- Email: [email protected]
The contribution of anticoagulants to platelet dysfunction with extracorporeal circulation
The contribution of anticoagulants to platelet dysfunction with extracorporeal circulation This investigation evaluates the effects of anticoagulants on platelet function. Fresh human blood from 40 nonmedicaled volunteers was anticoagulated with 4.3 units per milliliter heparin and lor acid-citrate-dextrose (ACD) solution 1:9. Retention of platelets from whole blood on glass beads was performed by the method of Bowie. Platelet retention of heparinized blood averaged 88.1 ± S.E. 1.5 per cent: ACD platelets averaged 24.6 ± S.E. 2.8 per cent. Platelet retention with citrate-phosphate-dextrose (CPD) and ethylenediaminetetraacetic acid (EDTA) yielded 26.0 ± S.E. 3.9 per cent and 19.1 ± S.E. 7.5 per cent, respectively. The addition of ACD to heparinized blood decreased platelet retention (19.7 ± S.E. 3.1 per cent). The addition of heparin to ACD or CPD blood did not alter the original decreased retention. Calcium added, even in excess, to blood containing heparin and ACD did not reverse the depressed retention (29.3 ± S.E. 4.6 per cent). The substitution of CPD gave similar results. With mixtures of separately collected ACD and heparininzed blood, depression of platelet retention was directly proportional to the amount of ACD blood present. Altering the pH of the ACD blood did not affect its depressed retention of platelets. Neutralizing heparinized blood 50 per cent with protamine or Polybrene also significantly depressed platelet retention 34.6 ± S.E. 5.8 per cent and 35.5 ± S.E. 4.0 per cent, respectively. Neither protamine nor Polybrene had any effect upon ACD blood. These data indicate that anticoagulants may play a significant role in the depressed platelet function observed during and following extracorporeal circulation.
Herbert W. Wallace, M.D., Helene Brooks, B.A. (by invitation), T. Peter Stein, Ph.D. (by invitation), and Nancy J. Zimmerman, B.A. (by invitation), Philadelphia, Pa.
.E/xtracorporeal circulation for either cardiopulmonary bypass or assisted membrane oxygenation has long been associated with a decrease in platelet count, which is independent of the type of prime and exceeds that expected from hemodilution alone.1, 2 Recent interest has turned to assessing possible alterations in platelet function resulting from extracorporeal circulation, since these may contribute to the hemostatic or thromboembolic complications following bypass.3 The present study was designed to evaluate the effects of anticoagulants and their antagonists on platelet function. From the Department of Surgery, School of Medicine of the University of Pennsylvania and The Graduate Hospital of the University of Pennsylvania, Philadelphia, Pa. 19146. Supported by National Institutes of Health Grant HL 15183. Read at the Fifty-sixth Annual Meeting of The American Association for Thoracic Surgery, Los Angeles, Calif., April 23, 24, and 25, 1976. Address for reprints: Herbert W. Wallace, M.D., Department of Surgery, The Graduate Hospital of the University of Pennsylvania, Philadelphia, Pa. 19146.
Methods and materials Platelet function amenable to study can arbitrarily be divided into three general processes: aggregation, release reaction, and adhesion. The platelet retention test of Bowie4 was chosen because we feel that conditions are well controlled and all three reactions usually occur during the platelet interaction with the glass bead surfaces. Fresh human blood was obtained by venipuncture from 40 volunteers who had taken no medication for at least one week. The blood was collected from the same donors for each set of experiments into plastic syringes containing enough heparin or acid-citrate-dextrose (ACD) solution (National Institutes of Health Formula A) to give a final concentration of 4.3 units per milliliter and/or 1:9 dilution, respectively. The glass bead columns were made by placing 2.6 Gm. of glass beads* (419 to 500 /A) into a 25 cm. length of polyvinyl tubing (I.D. 3 mm.) cut from a blood administration set, the end of which was plugged with a tiny piece of nylon mesh. The beads were tightly *Cataphore Division of Ferro Corp., Jackson, Miss.
735
7 36
The Journal of Thoracic and Cardiovascular Surgery
Wallace et al.
metal insert syringe ( l O c c ) tubing wire rack
TS~nylon
100 iHeparin
mesh
9.5 inches filled with glass beads
Fig. 1. A diagrammatic scheme of the method of measuring platelet retention.
packed by vibrating the plastic tubing vigorously with a vortex mixer. The excess tubing was cut off (final length: 23.75 cm.) so that the tip of the syringe fitted snugly against the glass bead column. All blood samples were evaluated within 2 hours after initial collection to ensure that any measured platelet dysfunction was not due to cellular deterioration. Following collection, a 1 ml. aliquot of blood was transferred to a 12 by 75 mm. plastic tube* containing 26 \ of 7 per cent solution of ethylenediaminetetraacetic acid (EDTA) for a control platelet count. The syringe was then placed on a slowly turning rotary mixer to ensure equal distribution of platelets in the sample, after which it was attached to the glass bead column and placed onto a Harvard syringe pumpt (Fig. 1). The blood was injected at a flow rate of 4.12 ml. per minute giving a contact time of 9.6 seconds. Effluent blood was collected in 1 ml. aliquots into plastic tubes containing 26 A. of a 7 per cent solution of EDTA to prevent platelet aggregation and clumping. The platelets were counted with a phase microscope by the method of Brecher and Cronkite.5 Platelet retention gradually increases in successive aliquots, reaching a maximum by the fourth or fifth milliliter collected. In these experiments, platelet retention is calculated from the average of these two aliquots to obtain the percentage of platelets retained on the glass bead column. Experiment A. Sixty-two samples of whole blood were collected: 25 with heparin, 29 with ACD, 4 with CPD (1:7), and 4 with EDTA (final concentration: 1.5 mg. per milliliter). These samples were evaluated for retention by the method previously described. Experiment B. To determine the effect of each anticoagulant upon the other in the same solution, ACD (1:9 dilution) was added to four samples of heparinized whole blood and heparin (4.3 units per milliliter) was added to eight ACD blood samples and two samples ♦Falcon, Oxnard, Calif. tHarvard Apparatus Co., Millis, Mass.
50 -
ACD
25
29
CPD
EDTA
4 Number
Fig. 2. The differences in platelet retention produced by anticoagulants. Heparin is the only one without a significant effect. anticoagulated with CPD before determining platelet retention. Experiment C. Since calcium ion is reported to be important in platelet retention,6- 7 two experiments were performed in which blood was separately anticoagulated with heparin and ACD from the same donors. Platelet retention was determined. Heparin was added to ACD blood and ACD was added to heparinized blood and retention determined. With these controls, heparin (4.3 units per milliliter) was added to the ACD blood to maintain anticoagulation, as well as ACD (1:9 dilution) added to the heparinized blood. To each of five aliquots of these mixtures was added 25, 50, 75, 100, and 200 per cent of the calculated quantity of calcium chloride necessary to completely counteract the calcium-binding capacity of the citrate present. Experiment D. The total volume of ACD is by necessity greater than that needed with heparin for similar anticoagulation. To evaluate the possible effect of dilution as well as to quantitate the effect of ACD upon platelets, a quantity of heparinized blood was collected from two volunteers and divided into the following experimental groups for platelet retention: heparin alone; heparinized blood to which one tenth the usual quantity of ACD was added; heparinized blood to which one half the usual quantity of ACD was added; heparinized blood to which one tenth the usual quantity of ACD was added followed by an excess of calcium;
Volume 72 Number 5 November, 1976
Platelet dysfunction with extracorporeal circulation
737
Table I. Effect of calcium on platelet retention of mixtures Mixture Heparin Heparin HEP -» HEP -» HEP -» HEP-* HEP -»
-> ACD ACD + 25% Ca ACD + 50% Ca ACD + 75% Ca ACD + 100% Ca ACD + 200% Ca
Platelet retention
<%)
90.6 12.5 10.5 8.0 41.6 41.0 51.3
Mixture ACD ACD ACD ACD ACD ACD ACD
-» -» -» -» -» -»
Heparin HEP + 25% Ca HEP + 50% Ca HEP + 75% Ca HEP + 100% Ca HEP + 200% Ca
Platelet retention (%) 13.9 15.4 6.0 14.7 28.9 30.4 34.0
Legend: Calcium, even in excess, does not reverse the effect of ACD on platelet retention.
heparinized blood to which one half the usual quantity of ACD was added followed by an excess of calcium; heparinized blood to which ACD in the usual quantity was added. Platelet retention of these samples was then evaluated. Experiment E. Blood anticoagulated with ACD and heparin was drawn separately from the same donors and mixed together in various proportions prior to the determination of platelet retention. To similar samples excess amounts of calcium chloride were added to surpass the binding of citrate prior to evaluating platelet retention. Experiment F. A series of five blood samples from each of 16 individuals was successively collected. The collecting syringes of these samples contained the following ratio of required anticoagulant: ACD 100 per cent; heparin 100 per cent; ACD 75 per cent and heparin 25 per cent; ACD 25 per cent and heparin 75 per cent; ACD 50 per cent and heparin 50 per cent. The retention ability of these samples was determined. Experiment G. By use of 1.5 M N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), a biologically inactive organic buffer, or 0.1 N sodium hydroxide, the pH of 12 ACD and 14 heparinized blood samples was adjusted in a range from 7.2 to 7.4 to evaluate the effect of hydrogen ion concentration upon platelet retention. Experiment H. In 17 experiments, heparinized and ACD blood samples were drawn from the same individuals and platelet retention determined. An aliquot of this same heparinized blood was titrated with protamine and hexadimethrine bromide (Polybrene) to determine the quantity necessary for coagulation. Fifty per cent of this amount was then added to aliquots of the remaining heparinized and ACD blood samples, and platelet retention was again determined. Results Experiment A. The effect of individual anticoagulants upon glass bead platelet retention is
graphically depicted in Fig. 2. Platelet retention of heparinized blood, as has been previously reported, averaged 88.1 ± S.E. 1.5 per cent. However, platelet retention with ACD-anticoagulated blood averaged only 24.6 ± S.E. 2.8 per cent. Platelet retention with CPD and EDTA yielded similar low values (26.0 ± S.E. 3.9 per cent and 19.1 ± S.E. 7.5 per cent, respectively). Experiment B. The addition of ACD to heparinized blood significantly decreased platelet retention (average: 19.7 ± S.E. 3.1 per cent) to levels similar to those seen for blood anticoagulated with ACD alone. On the other hand, the addition of heparin to ACD blood or CPD blood did not alter the original decreased retention (average: 16.8 ± S.E. 3.0 per cent and 34.9 ± S.E. 5.1 per cent, respectively). Experiment C. The results of the two experiments investigating the effect of calcium on platelet retention in the presence of anticoagulant mixtures are presented in Table I. The addition of up to 50 per cent of the calculated calcium ion necessary to counteract the citrate had no effect upon platelet retention, and only a minor recovery was observed even after the addition of 200 per cent of the calculated necessary calcium dose. Thus the addition of even a significant excess of calcium ion does not reverse the depressing effect of ACD upon platelet retention. Experiment D. The experiment to evaluate ACD dilution as well as ACD sensitivity of platelets reveals that the platelets are extremely sensitive to the presence of ACD, even in minimal amounts. The results are in Table II. These results clearly indicate the extreme sensitivity of platelets to the presence of ACD with a resulting depression in platelet retention comparable to the use of full-strength ACD. The addition of calcium did not reverse this process. The fractional amounts of ACD added were not large enough to alter pH significantly. Experiment E. When the platelet retention of mixtures of varying ratios of ACD and heparinized blood
7 38
The Journal of Thoracic and Cardiovascular Surgery
Wallace et al.
au -
70
I6j "8
7
-
50
9
>
I6~L,
joV^
30
io'
?2
Heporin ACD
10 90
1 20 80
1 30 70
40 60
I 50 50
60 40
I 70 30
80 20
l
90 10
l
100 0
Fig. 3. Mixtures of ACD and heparinized blood. With different ratios of ACD and heparinized blood, the platelet retention appears to be directly affected by the quantity of ACD present. Table II. Platelet retention (%) with minimal ACD
100 (0/100)
;,
(25/75)
HEP -> ACD
82 68
14 0
HEP -» y2 ACD
HEP -» '/,0ACD -> Ca
HEP -► 1/2 ACD -» Ca
35 18
11 16
26.5 1.4
(50/50)
50 (75/25)
A
14
13
14
l
28 33
Legend: As little as 10 per cent had as much effect on platelet retention as a full ACD quantity.
(100/0)
12
HEP -» lwACD
Heparin
16
Number
Fig. 4. Collection of blood in the presence of both ACD and heparin does not alter ACD effect on platelet retention.
was determined, the quantity of platelets retained in the glass bead column is inversely proportional to the concentration of ACD in the final mixture (Fig. 3). The addition of excess calcium ion to the mixtures did not significantly alter the platelet dysfunction apparently caused by the presence of ACD (Table III). These experiments again confirm the inability of the platelets to react normally in the presence of ACD anticoagulant. This dysfunction may be calcium dependent but cannot be reversed by the addition of calcium ion.
Table III. Platelet retention (%) of mixtures with calcium ACDIheparin
Mean S.E. N
20180
40160
60140
80/20
80 4.7 6
11.1 4.1 6
54.7 6.2 6
39.7 4.0 6
Legend: The addition of excess calcium ion had no effect on the platelet retention of various ACD/heparin ratios.
Experiment F. The results of the platelet retention data obtained from blood containing simultaneously heparin and ACD are almost identical to those found when similar mixtures are made from blood drawn separately into heparin or ACD. The data (Fig. 4) superimpose upon the curve of varying ACD/heparin ratios (Fig. 3). Thus the presence of heparin at no time prevents the refractory effect of ACD upon platelet function. Experiment G. Altering the pH of the ACD blood
Volume 72 Number 5 November, 1976
between 7.2 and 7.4 did not affect its depressed retention of platelets, nor did a change of pH influence platelet retention of heparinized blood. Experiment H. It is interesting to note that in the 17 experiments in which heparinized blood was 50 per cent neutralized with protamine or Polybrene, there was significant depression of platelet retention, with values of 34.6 ± S.E. 5.8 per cent and 35.5 ± S.E. 4.0, respectively. Neither protamine nor Polybrene had any effect upon the platelet retention of ACD blood. However, Polybrene with ACD caused platelet clumping on occasion. Discussion These studies are part of our ongoing effort to evaluate the factors associated with extracorporeal circulation which may influence platelets and thus affect coagulation and thromoembolism. We found it difficult to interpret the data obtained from patients undergoing heart surgery because of the variety of uncontrollable factors. McKenna and associates,8 in a recent study of patients before and after bypass, reported a preoperative platelet retention of 20 to 38 per cent as determined by the method of Bowie.4 This is in marked contrast to the usually reported normal values of 70 to 95 per cent. Since the preoperative platelets already respond abnormally, interpretation of the data is difficult and the effects of drugs, mechanical events, physiological responses, or other factors are impossible to differentiate. This reduced "normal" value most likely is not due to technical error but rather to the inherent problem of trying to control the platelet milieu in this group of patients, for we have found similarly low results in many of our patients preoperatively. This points out the serious difficulty in attempting to utilize uncontrolled patient material to determine true facts and mechanisms. In order to elucidate the details of the problem and perhaps understand mechanisms, we have decided to dissect out various aspects and study them individually in a more controllable but artificial system. We previously have shown that denatured plasma affects platelet aggregation.9 These present data demonstrate the influence of anticoagulants upon platelet function. It can be argued that the test employed is unphysiological, since platelets usually adhere to subendothelial or blood components. However, no standardized test for platelet adhesion to subendothelial tissues exists even though collagen is used to initiate aggregation in platelet-rich plasma. Collagen will often stimulate the release reaction and the released products (e.g., ADP) cause aggregation. Because the method of platelet re-
Platelet dysfunction with extracorporeal circulation
739
tention utilizing glass beads appears to induce the platelet to perform most, if not all, of its functions, we feel that this test is the best in vitro test to compare to the real-life situation. Thomson and his co-workers10 have shown that heparin increases platelet aggregation by ADP and adrenalin and enhances the platelet release reaction when tested in citrated platelet-rich plasma. In addition, utilizing cellophane membranes within a specially designed test cell, he noted that retention of platelets was increased when heparin was added to citrated whole blood. Utilizing glass cover slips, Mohammad and associates11 noted the platelets were more adhesive in citrated platelet-rich plasma than in the presence of EDTA or heparin. In the presence of citrate or heparin, they found platelet adhesion was greater in tests with polyethylene and polyvinyl chloride than with glass whereas, in the presence of EDTA, platelet adhesion decreased on these latter surfaces. Our present study confirms that anticoagulants do affect platelet function in vitro and probably in vivo. The presence of ACD apparently alters the ability of platelets to interact within a glass bead column. The partial neutralization of heparinized blood with protamine has a similar effect. O'Brien and associates7 felt that any influence of citrate upon platelet function is related to alterations in available calcium ions. However, in our study the addition of up to twice the calcium ion concentration that would be bound by the citrate did not influence this refractoriness to glass bead retention. We also found that as little as 10 per cent of the usual citrate concentration produced a marked depression of platelet retention. Thus human platelets in vitro are very sensitive to the presence of citrate. Although the binding of calcium ion may play a role, the effect is not reversible with the addition of calcium, even in excess. It is possible that citrate affects platelet membrane directly. Alterations in pH also have no significant influence. Coller and Zucker12 demonstrated that the mishandling of heparinized whole blood would temporarily influence platelet retention. Although we observed some of these same results with our heparinized samples, we are unable to prevent the depressing effect of ACD by alterations in method of blood handling. We are quite certain that mechanical factors do not influence these results. The assumption cannot be made that ACD absolutely affects platelet function in vivo, but the circumstantial evidence is strong. There is no evidence that any of the presently available in vitro tests necessarily reflect exactly what is occurring during platelet interactions with blood com-
740
The Journal of Thoracic and Cardiovascular Surgery
Wallace et al.
ponents and blood vessel walls. These arbitrary tests only reflect the behavior of a platelet in vitro and allow one to loosely interpret platelet physiology and biochemistry. Thus it is conceivable that platelets with an abnormal in vitro response may behave normally within its natural milieu. On the other hand, it must be remembered that a normal platelet count is not an indication of normal platelet function. While the concentration of heparin in these studies reflects those employed clinically, the concentrations of ACD utilized for most of the experiments are far in excess of the quantities usually found within the circulation except with massive transfusions. However, these concentrations are within the range used in banked blood and platelet concentrates or possibly found at the intravascular site of infusion (needle or catheter tip). Thus it is possible that damage to some platelets may occur before they reach the general circulation. In addition, the platelets' sensitivity to very small concentrations of citrate, which we have observed, may be clinically important. The interpretation of platelet function results is influenced not only by both controllable and uncontrollable factors, but also by the many variations in methodology employed by individual groups of investigators. This is a major difficulty in attempting to compare results from one study to another. Perhaps the time has come for investigators to agree upon a series of tests that they would all utilize. In addition, in those instances in which in vivo experimentation is required and clinical material is unsuitable, for whatever the reason, the problem of species variation in platelet and coagulation behavior must be dealt with. Again, it would be helpful to the progress of this area of investigation if a consensus could be reached of a specific limited number of experimental species which would be employed by all.
REFERENCES 1 Kendall, A. G., and Lowenstein, L.: Alterations in Blood Coagulation and Hemostasis During Extracorporeal Circulation, Can. Med. Assoc. J. 87: 786, 1962. 2 McKenzie, F. N., Dhall, D. P., Arfors, K. E., Nordlund, S., and Matheson, N. A.: Blood Platelet Behavior During and After Open Heart Surgery, Br. Med. J. 2: 796, 1969. 3 Solis, R. T., Beall, A. C , Jr., Noon, G. P., and De Bakey, M. E.: Platelet Aggregation: Effects of Cardiopulmonary Bypass, Chest 67: 558, 1975. 4 Bowie, E. J. W., Owen, C. A., Jr., Thompson, J. H., Jr., and Didisheim, P.: A Test of Platelet Adhesiveness, Mayo Clin. Proc. 44: 306, 1969. 5 Brecher, V., and Cronkite, E. T.: Morphology and
Enumeration of Human Blood Platelets, J. Appl. Physiol. 3: 365, 1950. 6 Salzman, E. W.: Measurement of Platelet Adhesiveness, J. Lab. Clin. Med. 62: 724, 1963. 7 O'Brien, J. R., Shoobridge, S. M., and Finch, W. J.: Comparison of the Effect of Heparin and Citrate on Platelet Aggregation, J. Clin. Pathol. 22: 28, 1969. 8 McKenna, R., Bachmann, F., Whittaker, B., Gilson, J. R., and Weinberg, J., Jr.: The Hemostatic Mechanism After Open-Heart Surgery. II. Frequency of Abnormal Platelet Functions During and After Extracorporeal Circulation, J. THORAC. CARDIOVASC. SURG. 70: 298,
1975.
9 Wallace, H. W., Liquori, E., Stein, T. P., and Brooks, H. B.: Denatured Plasma and Platelet Function, Trans. Am. Soc. Artif. Intern. Organs 21: 450, 1975. 10 Thomson, C , Forbes, C D., and Prentice, C. R. M.: The Potentiation of Platelet Aggregation and Adhesion by Heparin In Vitro and In Vivo, Clin. Sci. Molec. Med. 45:485, 1973. 11 Mohammad, S. F., Hardison, M. D., Glenn, C. H., Morton, B. D., Bolan, J. C , and Mason, R. G.: Adhesion of Human Blood Platelets to Glass and Polymer Surfaces, Haemostasis 3: 257, 1974. 12 Coller, B. S., andZucker, M. B.: Reversible Decrease in Platelet Retention by Glass Bead Columns (Adhesiveness) Induced by Disturbing the Blood, Proc. Soc. Exp. Biol. Med. 136: 769, 1971. Discussion DR. J O H N H.
KENNEDY
Cleveland, Ohio
In a fastidious study, Dr. Wallace and his co-workers have appropriately focused on the thrombocyte as the principal protagonist in the artificial surface-blood interface and have commented on the several factors which are operative in the perioperative period and will influence platelet behavior. In a current study of platelet function in dogs subjected to 5 hours of membrane oxygenation, we observed early aggregation and incipient release reaction with flipod extension. However, it should be added that anesthesia alone has produced these same changes. A proper understanding of platelet dynamics in patients and animals, therefore, depends upon an understanding of membrane dynamics of the platelet, and this is clearly a molecular problem. I certainly share Dr. Wallace's conviction that uniform methods of study of platelet function must be undertaken, and I would urge that we begin to consider the molecular aspects of platelet membrane function as well.
D R . C A R Y J.
LAMBERT
Dallas, Texas
We absolutely agree that platelet determinations and aggregational studies are meaningless under the circumstances described. On the other hand, we would like to point out that, with some exceptions, platelet usage or administration is not necessary in our program—the exceptions being
Volume 72 Number 5 November, 1976
patients with uremia, patients with platelet factor 3 disturbances best managed by dialysis, patients requiring long-term aspirin and persantum therapy, and patients with Marfan's syndrome. Postperfusion bleeding primarily occurs because of disturbances in one of four parameters. The first of these is faulty surgical technique; the second and primary cause in our hands, fibrinolysis; the third, factor deficiency; and the fourth, some combination thereof. Using a no-blood prime, accepting postoperative normovolemic anemia and only blood-component availability for the conduct of surgery, we have based our program upon an exacting and reproducible method of hemostatic assessment such that we can determine whether there is a faulty disturbance in terms of heparin, circulating anticoagulants, fibrinolysis, or factor deficiency. Consequently, we do not use whole blood of any sort at any time. For cardiovascular surgery, vascular surgery, or thoracic surgery, we only set up 3 units of pack cells. In the last 647 patients, only 4, an incidence of less than 1 per cent, have required reoperation—3 of those being for tamponade. We do not use platelets nor advocate the need for platelets. We have noted that bleeding primarily occurs within the first 3 hours following bypass and, with appropriate diagnosis and treatment, has averaged less than 350 c.c. the first 8 hours. Blood volume determinations despite "anemia" have tended to be normal, and hemoglobin levels have reverted to normal within 3 weeks. Faulty wound healing, altered cardiac function, and tissue delivery of oxygen have not been problems. DR. D E N T O N A . C O O L E Y Houston, Texas
Dr. Wallace has brought some very important observations to our attention. In the preamble to his paper, he mentioned tests other than platelet adhesiveness which are useful in assessing platelet function. According to hematologists at my hospital, the study which he describes using glass beads is not entirely reliable because of the variation and size of the beads and other variables. Our laboratory uses an aggremometer to assess platelet cohesiveness or the ability of the platelets to aggregate. In studies on our patients undergoing extracorporeal circulation, our staff was unable to determine the presence of any dysfunction whatsoever of the platelets and could only demonstrate a decrease in concentration. It may be of interest, too, that this decrease began before the patients received heparin. We noted a decrease in platelets of the order of 20 to 30 per cent just after the sternotomy was made. Thus the sternotomy alone accounts for a significant reduction in the platelet concentration. During the pumping procedure and after heparin had been administered, a progressive drop in the platelet count occurs. Immediately after pump and before heparin was reversed, the platelet count reached its lowest level. Twenty-four hours
Platelet dysfunction with extracorporeal
circulation
14 1
later the platelet concentration returned to normal, but the platelets were functioning normally. We undertook a study about a year ago to sequester blood (platelets, erythrocytes, and leukocytes) by applying leg tourniquets during open-heart surgery to determine if any beneficial effect resulted in blood cytology. We could not demonstrate any difference in a control group compared to an experimental group. We were able to sequester about 20 per cent of the total blood volume during these surgical procedures. The studies presented by Dr. Wallace today are quite significant, because, again, it incriminates ACD-treated bank blood in cardiovascular surgery. We believe that bank blood actually increases the incidence of postoperative bleeding. It may also account for the fact that in some 300 patients who were Jehovah's Witnesses and who have undergone extracorporeal circulation at our hospital, the incidence of postoperative bleeding has been substantially less than in routine cases. D R . W A L L A C E (Closing) I wish to thank the three discussers for their remarks. Dr. Kennedy, I agree that platelet membrane function is very important to study, and I think that this is the area on which one now needs to begin to concentrate. However, before reaching that point, I think one needs to have some idea in what directions to go and, therefore, what some of the problems may be. Again, it is hard to know what a normal function would be of a platelet once we take it out of its mileau and start isolating some of the membrane components. Dr. Lambert, I also agree with you that there is absolutely no substitute for good surgical technique. Dr. Cooley, what you point out about the differences in tests is exactly what I was saying at the end of my talk—that what we need is to have some standardization of the tests so we can compare results between groups. I have done the aggregation tests with the same electronic equipment, I am sure, that you and everybody else uses. I chose this particular test because it was very well controlled. The bead sizes and the number of beads were well controlled. We used exactly 2.6 Gm. of bead, and each bead was within a very few microns in size. The size of the tubing and the rate at which the blood goes through the tubing was well controlled. For all of these reasons, plus the fact that the results are reproducible, we utilized this test instead of aggregation, which we found to be less well controlled. Aggregation also has some differences if you evaluate the various anticoagulants. The fact that you do not see any differences in aggregation at the end of perfusion is not surprising to me and points out again the reason why patients supported by bypass are sometimes poor models to use to try to define the problem. It may be that, with the drop in platelet levels, those platelets which were damaged were removed from the circulation and those that are left may still be functioning fairly well.
Herbert W. Wallace, M.D., Helene Brooks, B.A. (by invitation), T. Peter Stein, Ph.D. (by invitation), and Nancy J. Zimmerman, B.A. (by invitation), Philadelphia, Pa.
.E/xtracorporeal circulation for either cardiopulmonary bypass or assisted membrane oxygenation has long been associated with a decrease in platelet count, which is independent of the type of prime and exceeds that expected from hemodilution alone.1, 2 Recent interest has turned to assessing possible alterations in platelet function resulting from extracorporeal circulation, since these may contribute to the hemostatic or thromboembolic complications following bypass.3 The present study was designed to evaluate the effects of anticoagulants and their antagonists on platelet function. From the Department of Surgery, School of Medicine of the University of Pennsylvania and The Graduate Hospital of the University of Pennsylvania, Philadelphia, Pa. 19146. Supported by National Institutes of Health Grant HL 15183. Read at the Fifty-sixth Annual Meeting of The American Association for Thoracic Surgery, Los Angeles, Calif., April 23, 24, and 25, 1976. Address for reprints: Herbert W. Wallace, M.D., Department of Surgery, The Graduate Hospital of the University of Pennsylvania, Philadelphia, Pa. 19146.
Methods and materials Platelet function amenable to study can arbitrarily be divided into three general processes: aggregation, release reaction, and adhesion. The platelet retention test of Bowie4 was chosen because we feel that conditions are well controlled and all three reactions usually occur during the platelet interaction with the glass bead surfaces. Fresh human blood was obtained by venipuncture from 40 volunteers who had taken no medication for at least one week. The blood was collected from the same donors for each set of experiments into plastic syringes containing enough heparin or acid-citrate-dextrose (ACD) solution (National Institutes of Health Formula A) to give a final concentration of 4.3 units per milliliter and/or 1:9 dilution, respectively. The glass bead columns were made by placing 2.6 Gm. of glass beads* (419 to 500 /A) into a 25 cm. length of polyvinyl tubing (I.D. 3 mm.) cut from a blood administration set, the end of which was plugged with a tiny piece of nylon mesh. The beads were tightly *Cataphore Division of Ferro Corp., Jackson, Miss.
735
7 36
The Journal of Thoracic and Cardiovascular Surgery
Wallace et al.
metal insert syringe ( l O c c ) tubing wire rack
TS~nylon
100 iHeparin
mesh
9.5 inches filled with glass beads
Fig. 1. A diagrammatic scheme of the method of measuring platelet retention.
packed by vibrating the plastic tubing vigorously with a vortex mixer. The excess tubing was cut off (final length: 23.75 cm.) so that the tip of the syringe fitted snugly against the glass bead column. All blood samples were evaluated within 2 hours after initial collection to ensure that any measured platelet dysfunction was not due to cellular deterioration. Following collection, a 1 ml. aliquot of blood was transferred to a 12 by 75 mm. plastic tube* containing 26 \ of 7 per cent solution of ethylenediaminetetraacetic acid (EDTA) for a control platelet count. The syringe was then placed on a slowly turning rotary mixer to ensure equal distribution of platelets in the sample, after which it was attached to the glass bead column and placed onto a Harvard syringe pumpt (Fig. 1). The blood was injected at a flow rate of 4.12 ml. per minute giving a contact time of 9.6 seconds. Effluent blood was collected in 1 ml. aliquots into plastic tubes containing 26 A. of a 7 per cent solution of EDTA to prevent platelet aggregation and clumping. The platelets were counted with a phase microscope by the method of Brecher and Cronkite.5 Platelet retention gradually increases in successive aliquots, reaching a maximum by the fourth or fifth milliliter collected. In these experiments, platelet retention is calculated from the average of these two aliquots to obtain the percentage of platelets retained on the glass bead column. Experiment A. Sixty-two samples of whole blood were collected: 25 with heparin, 29 with ACD, 4 with CPD (1:7), and 4 with EDTA (final concentration: 1.5 mg. per milliliter). These samples were evaluated for retention by the method previously described. Experiment B. To determine the effect of each anticoagulant upon the other in the same solution, ACD (1:9 dilution) was added to four samples of heparinized whole blood and heparin (4.3 units per milliliter) was added to eight ACD blood samples and two samples ♦Falcon, Oxnard, Calif. tHarvard Apparatus Co., Millis, Mass.
50 -
ACD
25
29
CPD
EDTA
4 Number
Fig. 2. The differences in platelet retention produced by anticoagulants. Heparin is the only one without a significant effect. anticoagulated with CPD before determining platelet retention. Experiment C. Since calcium ion is reported to be important in platelet retention,6- 7 two experiments were performed in which blood was separately anticoagulated with heparin and ACD from the same donors. Platelet retention was determined. Heparin was added to ACD blood and ACD was added to heparinized blood and retention determined. With these controls, heparin (4.3 units per milliliter) was added to the ACD blood to maintain anticoagulation, as well as ACD (1:9 dilution) added to the heparinized blood. To each of five aliquots of these mixtures was added 25, 50, 75, 100, and 200 per cent of the calculated quantity of calcium chloride necessary to completely counteract the calcium-binding capacity of the citrate present. Experiment D. The total volume of ACD is by necessity greater than that needed with heparin for similar anticoagulation. To evaluate the possible effect of dilution as well as to quantitate the effect of ACD upon platelets, a quantity of heparinized blood was collected from two volunteers and divided into the following experimental groups for platelet retention: heparin alone; heparinized blood to which one tenth the usual quantity of ACD was added; heparinized blood to which one half the usual quantity of ACD was added; heparinized blood to which one tenth the usual quantity of ACD was added followed by an excess of calcium;
Volume 72 Number 5 November, 1976
Platelet dysfunction with extracorporeal circulation
737
Table I. Effect of calcium on platelet retention of mixtures Mixture Heparin Heparin HEP -» HEP -» HEP -» HEP-* HEP -»
-> ACD ACD + 25% Ca ACD + 50% Ca ACD + 75% Ca ACD + 100% Ca ACD + 200% Ca
Platelet retention
<%)
90.6 12.5 10.5 8.0 41.6 41.0 51.3
Mixture ACD ACD ACD ACD ACD ACD ACD
-» -» -» -» -» -»
Heparin HEP + 25% Ca HEP + 50% Ca HEP + 75% Ca HEP + 100% Ca HEP + 200% Ca
Platelet retention (%) 13.9 15.4 6.0 14.7 28.9 30.4 34.0
Legend: Calcium, even in excess, does not reverse the effect of ACD on platelet retention.
heparinized blood to which one half the usual quantity of ACD was added followed by an excess of calcium; heparinized blood to which ACD in the usual quantity was added. Platelet retention of these samples was then evaluated. Experiment E. Blood anticoagulated with ACD and heparin was drawn separately from the same donors and mixed together in various proportions prior to the determination of platelet retention. To similar samples excess amounts of calcium chloride were added to surpass the binding of citrate prior to evaluating platelet retention. Experiment F. A series of five blood samples from each of 16 individuals was successively collected. The collecting syringes of these samples contained the following ratio of required anticoagulant: ACD 100 per cent; heparin 100 per cent; ACD 75 per cent and heparin 25 per cent; ACD 25 per cent and heparin 75 per cent; ACD 50 per cent and heparin 50 per cent. The retention ability of these samples was determined. Experiment G. By use of 1.5 M N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), a biologically inactive organic buffer, or 0.1 N sodium hydroxide, the pH of 12 ACD and 14 heparinized blood samples was adjusted in a range from 7.2 to 7.4 to evaluate the effect of hydrogen ion concentration upon platelet retention. Experiment H. In 17 experiments, heparinized and ACD blood samples were drawn from the same individuals and platelet retention determined. An aliquot of this same heparinized blood was titrated with protamine and hexadimethrine bromide (Polybrene) to determine the quantity necessary for coagulation. Fifty per cent of this amount was then added to aliquots of the remaining heparinized and ACD blood samples, and platelet retention was again determined. Results Experiment A. The effect of individual anticoagulants upon glass bead platelet retention is
graphically depicted in Fig. 2. Platelet retention of heparinized blood, as has been previously reported, averaged 88.1 ± S.E. 1.5 per cent. However, platelet retention with ACD-anticoagulated blood averaged only 24.6 ± S.E. 2.8 per cent. Platelet retention with CPD and EDTA yielded similar low values (26.0 ± S.E. 3.9 per cent and 19.1 ± S.E. 7.5 per cent, respectively). Experiment B. The addition of ACD to heparinized blood significantly decreased platelet retention (average: 19.7 ± S.E. 3.1 per cent) to levels similar to those seen for blood anticoagulated with ACD alone. On the other hand, the addition of heparin to ACD blood or CPD blood did not alter the original decreased retention (average: 16.8 ± S.E. 3.0 per cent and 34.9 ± S.E. 5.1 per cent, respectively). Experiment C. The results of the two experiments investigating the effect of calcium on platelet retention in the presence of anticoagulant mixtures are presented in Table I. The addition of up to 50 per cent of the calculated calcium ion necessary to counteract the citrate had no effect upon platelet retention, and only a minor recovery was observed even after the addition of 200 per cent of the calculated necessary calcium dose. Thus the addition of even a significant excess of calcium ion does not reverse the depressing effect of ACD upon platelet retention. Experiment D. The experiment to evaluate ACD dilution as well as ACD sensitivity of platelets reveals that the platelets are extremely sensitive to the presence of ACD, even in minimal amounts. The results are in Table II. These results clearly indicate the extreme sensitivity of platelets to the presence of ACD with a resulting depression in platelet retention comparable to the use of full-strength ACD. The addition of calcium did not reverse this process. The fractional amounts of ACD added were not large enough to alter pH significantly. Experiment E. When the platelet retention of mixtures of varying ratios of ACD and heparinized blood
7 38
The Journal of Thoracic and Cardiovascular Surgery
Wallace et al.
au -
70
I6j "8
7
-
50
9
>
I6~L,
joV^
30
io'
?2
Heporin ACD
10 90
1 20 80
1 30 70
40 60
I 50 50
60 40
I 70 30
80 20
l
90 10
l
100 0
Fig. 3. Mixtures of ACD and heparinized blood. With different ratios of ACD and heparinized blood, the platelet retention appears to be directly affected by the quantity of ACD present. Table II. Platelet retention (%) with minimal ACD
100 (0/100)
;,
(25/75)
HEP -> ACD
82 68
14 0
HEP -» y2 ACD
HEP -» '/,0ACD -> Ca
HEP -► 1/2 ACD -» Ca
35 18
11 16
26.5 1.4
(50/50)
50 (75/25)
A
14
13
14
l
28 33
Legend: As little as 10 per cent had as much effect on platelet retention as a full ACD quantity.
(100/0)
12
HEP -» lwACD
Heparin
16
Number
Fig. 4. Collection of blood in the presence of both ACD and heparin does not alter ACD effect on platelet retention.
was determined, the quantity of platelets retained in the glass bead column is inversely proportional to the concentration of ACD in the final mixture (Fig. 3). The addition of excess calcium ion to the mixtures did not significantly alter the platelet dysfunction apparently caused by the presence of ACD (Table III). These experiments again confirm the inability of the platelets to react normally in the presence of ACD anticoagulant. This dysfunction may be calcium dependent but cannot be reversed by the addition of calcium ion.
Table III. Platelet retention (%) of mixtures with calcium ACDIheparin
Mean S.E. N
20180
40160
60140
80/20
80 4.7 6
11.1 4.1 6
54.7 6.2 6
39.7 4.0 6
Legend: The addition of excess calcium ion had no effect on the platelet retention of various ACD/heparin ratios.
Experiment F. The results of the platelet retention data obtained from blood containing simultaneously heparin and ACD are almost identical to those found when similar mixtures are made from blood drawn separately into heparin or ACD. The data (Fig. 4) superimpose upon the curve of varying ACD/heparin ratios (Fig. 3). Thus the presence of heparin at no time prevents the refractory effect of ACD upon platelet function. Experiment G. Altering the pH of the ACD blood
Volume 72 Number 5 November, 1976
between 7.2 and 7.4 did not affect its depressed retention of platelets, nor did a change of pH influence platelet retention of heparinized blood. Experiment H. It is interesting to note that in the 17 experiments in which heparinized blood was 50 per cent neutralized with protamine or Polybrene, there was significant depression of platelet retention, with values of 34.6 ± S.E. 5.8 per cent and 35.5 ± S.E. 4.0, respectively. Neither protamine nor Polybrene had any effect upon the platelet retention of ACD blood. However, Polybrene with ACD caused platelet clumping on occasion. Discussion These studies are part of our ongoing effort to evaluate the factors associated with extracorporeal circulation which may influence platelets and thus affect coagulation and thromoembolism. We found it difficult to interpret the data obtained from patients undergoing heart surgery because of the variety of uncontrollable factors. McKenna and associates,8 in a recent study of patients before and after bypass, reported a preoperative platelet retention of 20 to 38 per cent as determined by the method of Bowie.4 This is in marked contrast to the usually reported normal values of 70 to 95 per cent. Since the preoperative platelets already respond abnormally, interpretation of the data is difficult and the effects of drugs, mechanical events, physiological responses, or other factors are impossible to differentiate. This reduced "normal" value most likely is not due to technical error but rather to the inherent problem of trying to control the platelet milieu in this group of patients, for we have found similarly low results in many of our patients preoperatively. This points out the serious difficulty in attempting to utilize uncontrolled patient material to determine true facts and mechanisms. In order to elucidate the details of the problem and perhaps understand mechanisms, we have decided to dissect out various aspects and study them individually in a more controllable but artificial system. We previously have shown that denatured plasma affects platelet aggregation.9 These present data demonstrate the influence of anticoagulants upon platelet function. It can be argued that the test employed is unphysiological, since platelets usually adhere to subendothelial or blood components. However, no standardized test for platelet adhesion to subendothelial tissues exists even though collagen is used to initiate aggregation in platelet-rich plasma. Collagen will often stimulate the release reaction and the released products (e.g., ADP) cause aggregation. Because the method of platelet re-
Platelet dysfunction with extracorporeal circulation
739
tention utilizing glass beads appears to induce the platelet to perform most, if not all, of its functions, we feel that this test is the best in vitro test to compare to the real-life situation. Thomson and his co-workers10 have shown that heparin increases platelet aggregation by ADP and adrenalin and enhances the platelet release reaction when tested in citrated platelet-rich plasma. In addition, utilizing cellophane membranes within a specially designed test cell, he noted that retention of platelets was increased when heparin was added to citrated whole blood. Utilizing glass cover slips, Mohammad and associates11 noted the platelets were more adhesive in citrated platelet-rich plasma than in the presence of EDTA or heparin. In the presence of citrate or heparin, they found platelet adhesion was greater in tests with polyethylene and polyvinyl chloride than with glass whereas, in the presence of EDTA, platelet adhesion decreased on these latter surfaces. Our present study confirms that anticoagulants do affect platelet function in vitro and probably in vivo. The presence of ACD apparently alters the ability of platelets to interact within a glass bead column. The partial neutralization of heparinized blood with protamine has a similar effect. O'Brien and associates7 felt that any influence of citrate upon platelet function is related to alterations in available calcium ions. However, in our study the addition of up to twice the calcium ion concentration that would be bound by the citrate did not influence this refractoriness to glass bead retention. We also found that as little as 10 per cent of the usual citrate concentration produced a marked depression of platelet retention. Thus human platelets in vitro are very sensitive to the presence of citrate. Although the binding of calcium ion may play a role, the effect is not reversible with the addition of calcium, even in excess. It is possible that citrate affects platelet membrane directly. Alterations in pH also have no significant influence. Coller and Zucker12 demonstrated that the mishandling of heparinized whole blood would temporarily influence platelet retention. Although we observed some of these same results with our heparinized samples, we are unable to prevent the depressing effect of ACD by alterations in method of blood handling. We are quite certain that mechanical factors do not influence these results. The assumption cannot be made that ACD absolutely affects platelet function in vivo, but the circumstantial evidence is strong. There is no evidence that any of the presently available in vitro tests necessarily reflect exactly what is occurring during platelet interactions with blood com-
740
The Journal of Thoracic and Cardiovascular Surgery
Wallace et al.
ponents and blood vessel walls. These arbitrary tests only reflect the behavior of a platelet in vitro and allow one to loosely interpret platelet physiology and biochemistry. Thus it is conceivable that platelets with an abnormal in vitro response may behave normally within its natural milieu. On the other hand, it must be remembered that a normal platelet count is not an indication of normal platelet function. While the concentration of heparin in these studies reflects those employed clinically, the concentrations of ACD utilized for most of the experiments are far in excess of the quantities usually found within the circulation except with massive transfusions. However, these concentrations are within the range used in banked blood and platelet concentrates or possibly found at the intravascular site of infusion (needle or catheter tip). Thus it is possible that damage to some platelets may occur before they reach the general circulation. In addition, the platelets' sensitivity to very small concentrations of citrate, which we have observed, may be clinically important. The interpretation of platelet function results is influenced not only by both controllable and uncontrollable factors, but also by the many variations in methodology employed by individual groups of investigators. This is a major difficulty in attempting to compare results from one study to another. Perhaps the time has come for investigators to agree upon a series of tests that they would all utilize. In addition, in those instances in which in vivo experimentation is required and clinical material is unsuitable, for whatever the reason, the problem of species variation in platelet and coagulation behavior must be dealt with. Again, it would be helpful to the progress of this area of investigation if a consensus could be reached of a specific limited number of experimental species which would be employed by all.
REFERENCES 1 Kendall, A. G., and Lowenstein, L.: Alterations in Blood Coagulation and Hemostasis During Extracorporeal Circulation, Can. Med. Assoc. J. 87: 786, 1962. 2 McKenzie, F. N., Dhall, D. P., Arfors, K. E., Nordlund, S., and Matheson, N. A.: Blood Platelet Behavior During and After Open Heart Surgery, Br. Med. J. 2: 796, 1969. 3 Solis, R. T., Beall, A. C , Jr., Noon, G. P., and De Bakey, M. E.: Platelet Aggregation: Effects of Cardiopulmonary Bypass, Chest 67: 558, 1975. 4 Bowie, E. J. W., Owen, C. A., Jr., Thompson, J. H., Jr., and Didisheim, P.: A Test of Platelet Adhesiveness, Mayo Clin. Proc. 44: 306, 1969. 5 Brecher, V., and Cronkite, E. T.: Morphology and
Enumeration of Human Blood Platelets, J. Appl. Physiol. 3: 365, 1950. 6 Salzman, E. W.: Measurement of Platelet Adhesiveness, J. Lab. Clin. Med. 62: 724, 1963. 7 O'Brien, J. R., Shoobridge, S. M., and Finch, W. J.: Comparison of the Effect of Heparin and Citrate on Platelet Aggregation, J. Clin. Pathol. 22: 28, 1969. 8 McKenna, R., Bachmann, F., Whittaker, B., Gilson, J. R., and Weinberg, J., Jr.: The Hemostatic Mechanism After Open-Heart Surgery. II. Frequency of Abnormal Platelet Functions During and After Extracorporeal Circulation, J. THORAC. CARDIOVASC. SURG. 70: 298,
1975.
9 Wallace, H. W., Liquori, E., Stein, T. P., and Brooks, H. B.: Denatured Plasma and Platelet Function, Trans. Am. Soc. Artif. Intern. Organs 21: 450, 1975. 10 Thomson, C , Forbes, C D., and Prentice, C. R. M.: The Potentiation of Platelet Aggregation and Adhesion by Heparin In Vitro and In Vivo, Clin. Sci. Molec. Med. 45:485, 1973. 11 Mohammad, S. F., Hardison, M. D., Glenn, C. H., Morton, B. D., Bolan, J. C , and Mason, R. G.: Adhesion of Human Blood Platelets to Glass and Polymer Surfaces, Haemostasis 3: 257, 1974. 12 Coller, B. S., andZucker, M. B.: Reversible Decrease in Platelet Retention by Glass Bead Columns (Adhesiveness) Induced by Disturbing the Blood, Proc. Soc. Exp. Biol. Med. 136: 769, 1971. Discussion DR. J O H N H.
KENNEDY
Cleveland, Ohio
In a fastidious study, Dr. Wallace and his co-workers have appropriately focused on the thrombocyte as the principal protagonist in the artificial surface-blood interface and have commented on the several factors which are operative in the perioperative period and will influence platelet behavior. In a current study of platelet function in dogs subjected to 5 hours of membrane oxygenation, we observed early aggregation and incipient release reaction with flipod extension. However, it should be added that anesthesia alone has produced these same changes. A proper understanding of platelet dynamics in patients and animals, therefore, depends upon an understanding of membrane dynamics of the platelet, and this is clearly a molecular problem. I certainly share Dr. Wallace's conviction that uniform methods of study of platelet function must be undertaken, and I would urge that we begin to consider the molecular aspects of platelet membrane function as well.
D R . C A R Y J.
LAMBERT
Dallas, Texas
We absolutely agree that platelet determinations and aggregational studies are meaningless under the circumstances described. On the other hand, we would like to point out that, with some exceptions, platelet usage or administration is not necessary in our program—the exceptions being
Volume 72 Number 5 November, 1976
patients with uremia, patients with platelet factor 3 disturbances best managed by dialysis, patients requiring long-term aspirin and persantum therapy, and patients with Marfan's syndrome. Postperfusion bleeding primarily occurs because of disturbances in one of four parameters. The first of these is faulty surgical technique; the second and primary cause in our hands, fibrinolysis; the third, factor deficiency; and the fourth, some combination thereof. Using a no-blood prime, accepting postoperative normovolemic anemia and only blood-component availability for the conduct of surgery, we have based our program upon an exacting and reproducible method of hemostatic assessment such that we can determine whether there is a faulty disturbance in terms of heparin, circulating anticoagulants, fibrinolysis, or factor deficiency. Consequently, we do not use whole blood of any sort at any time. For cardiovascular surgery, vascular surgery, or thoracic surgery, we only set up 3 units of pack cells. In the last 647 patients, only 4, an incidence of less than 1 per cent, have required reoperation—3 of those being for tamponade. We do not use platelets nor advocate the need for platelets. We have noted that bleeding primarily occurs within the first 3 hours following bypass and, with appropriate diagnosis and treatment, has averaged less than 350 c.c. the first 8 hours. Blood volume determinations despite "anemia" have tended to be normal, and hemoglobin levels have reverted to normal within 3 weeks. Faulty wound healing, altered cardiac function, and tissue delivery of oxygen have not been problems. DR. D E N T O N A . C O O L E Y Houston, Texas
Dr. Wallace has brought some very important observations to our attention. In the preamble to his paper, he mentioned tests other than platelet adhesiveness which are useful in assessing platelet function. According to hematologists at my hospital, the study which he describes using glass beads is not entirely reliable because of the variation and size of the beads and other variables. Our laboratory uses an aggremometer to assess platelet cohesiveness or the ability of the platelets to aggregate. In studies on our patients undergoing extracorporeal circulation, our staff was unable to determine the presence of any dysfunction whatsoever of the platelets and could only demonstrate a decrease in concentration. It may be of interest, too, that this decrease began before the patients received heparin. We noted a decrease in platelets of the order of 20 to 30 per cent just after the sternotomy was made. Thus the sternotomy alone accounts for a significant reduction in the platelet concentration. During the pumping procedure and after heparin had been administered, a progressive drop in the platelet count occurs. Immediately after pump and before heparin was reversed, the platelet count reached its lowest level. Twenty-four hours
Platelet dysfunction with extracorporeal
circulation
14 1
later the platelet concentration returned to normal, but the platelets were functioning normally. We undertook a study about a year ago to sequester blood (platelets, erythrocytes, and leukocytes) by applying leg tourniquets during open-heart surgery to determine if any beneficial effect resulted in blood cytology. We could not demonstrate any difference in a control group compared to an experimental group. We were able to sequester about 20 per cent of the total blood volume during these surgical procedures. The studies presented by Dr. Wallace today are quite significant, because, again, it incriminates ACD-treated bank blood in cardiovascular surgery. We believe that bank blood actually increases the incidence of postoperative bleeding. It may also account for the fact that in some 300 patients who were Jehovah's Witnesses and who have undergone extracorporeal circulation at our hospital, the incidence of postoperative bleeding has been substantially less than in routine cases. D R . W A L L A C E (Closing) I wish to thank the three discussers for their remarks. Dr. Kennedy, I agree that platelet membrane function is very important to study, and I think that this is the area on which one now needs to begin to concentrate. However, before reaching that point, I think one needs to have some idea in what directions to go and, therefore, what some of the problems may be. Again, it is hard to know what a normal function would be of a platelet once we take it out of its mileau and start isolating some of the membrane components. Dr. Lambert, I also agree with you that there is absolutely no substitute for good surgical technique. Dr. Cooley, what you point out about the differences in tests is exactly what I was saying at the end of my talk—that what we need is to have some standardization of the tests so we can compare results between groups. I have done the aggregation tests with the same electronic equipment, I am sure, that you and everybody else uses. I chose this particular test because it was very well controlled. The bead sizes and the number of beads were well controlled. We used exactly 2.6 Gm. of bead, and each bead was within a very few microns in size. The size of the tubing and the rate at which the blood goes through the tubing was well controlled. For all of these reasons, plus the fact that the results are reproducible, we utilized this test instead of aggregation, which we found to be less well controlled. Aggregation also has some differences if you evaluate the various anticoagulants. The fact that you do not see any differences in aggregation at the end of perfusion is not surprising to me and points out again the reason why patients supported by bypass are sometimes poor models to use to try to define the problem. It may be that, with the drop in platelet levels, those platelets which were damaged were removed from the circulation and those that are left may still be functioning fairly well.