US3814687A - Blood platelet separation - Google Patents

Abstract

A METHOD OF SEPARATING PLATELETS FROM WHOLE BLOOD OR BLOOD CONSTITUENT BY SEPARATION FRACTIONATION IS DISCLOSED IN WHICH A COMBINATION OF REAGENTS INCLUDING FORMALDEHYDE OR OTHER SUITABLE BLOOD FIXATIVE AGENT AND A SEPARATION DEXTRAN POLYMER OR OTHER SUITABLE SEPARATION AGENT ARE ADDED TO THE WHOLE BLOOD FOR SEPARATING A PLATELETCONTAINING FRACTION THEREOF AND THE RESULTING MIXTURE ALLOWED TO SEPARATE LEAVING A LAYERED STRUCTURE FROM WHICH THE DESIRED FRACTION MAY BE EASILY RECOVERED.

Description

United States Patent 6,814,687 BLOOD PLATELET SEPARATION William D. Ellis, Bloomington, and Bruce D. Mulvaney,

Hopkins, Minn., assignors to Honeywell Inc., Minneapolis, Minn. No Drawing. Filed June '1, 1973, Ser. No. 365,884

Int. Cl. B01d 21/00 U.S. Cl. 210-49 20 Claims ABSTRACT OF THE DISCLOSURE A method of separating platelets from whole blood or blood constituent by separation fractionation is disclosed in which a combination of reagents including formaldehyde or other suitable blood fixative agent and a separation dextran polymer or other suitable separation agent are added to the whole blood for separating a plateletcontaining fraction thereof and the resulting mixture allowed to separate leaving a layered structure from which the desired fraction may be easily recovered.

BACKGROUND OF THE INVENTION Field of the invention The invention relates generally to the separation of the constituents of whole blood and, more particularly, to the separation of platelets from other blood constituents.

Description of the prior art Normal whole blood consists of 45 to 65 percent plasma which is a fluid constituent and the remainder is made up of cellular components. The cellular fraction in each cubic millimeter of whole blood includes approximately four million to 6.5 million red cells (erythrocytes), 150,- 000 to 400,000 platelets and 4,000 to 10,000 white cells (leukocytes).

Several important laboratory tests involving blood which are quite indicative of the state of health of the patient involved are greatly facilitated by the separation of various blood constituents so that certain of these may be independently counted or evaluated. One of the most important of these is known as the leukocyte differential count in which a representative sample of white blood cells (of which there are a number of distinct kinds) is counted to determine the relative numbers of the several kinds of these cells which are present in the sample. An accurate determination of the relative numbers of the various kinds of leukocytes present in an individuals blood is extremely important in the diagnosis and evaluation of many conditions.

Because of the importance of this and other tests and because the results of any differential count are directly related to the quality of the sample utilized for the count, it is very important that the sample be one in which the other constituents of the blood which might be mistaken for leukocytes are removed therefrom as much as possible. Also, the relative numbers of the various types of leukocytes must as closely as possible represent the relative numbers which exist in the whole blood from which the sample is derived. The need for a relatively pure, representative sample has also greatly increased with the advent of automated systems for leukocyte difierential counting in helping to eliminate errors resulting from improper differentiation of leukocyte images from erythrocytes (red blood cells), platelets or other debris in the sample. Such a sample also facilities recognition of distinct types of leukocytes. Thus, for such a test, whether it be accomplished by the eye of a laboratory technician or through an automated counting device could be made a great deal more accurate and faster if a sample could be prepared consisting solely of concentrated leukocytes.

A number of methods have been used in the prior art to attempt to isolate a concentration of leukocytes from whole blood. These include the selective destruction of erythrocytes, natural sedimentation, rolueaux promoted acceleration of sedimentation, simple centrifugation, leukoadhesion, the use of phagocytic properties and density gradient methods. These methods are discussed in detail by J. H. Cutts in Cell Separation Methods in Hematology, Academic Press, New York, 1970 and by R. L. Walford in Leukocyte Antigens and Antibodies, Grune and Stratton, New York, 1960.

While these enumerated methods have been found to be at least partially acceptable in obtaining a concentrate of leukocytes, each suffers from one or more disadvantages as procedures to prepare such a sample for a diiferential count. Thus, the selective destruction of erythrocytes often leads to unacceptable changes in white cell morphology (outward cell appearance) which leads to errors caused by misidentification of such cells in the differential count. Sedimentation, including accelerated techniques, take a long time to prepare (at least thirty minutes) and can result in a concentrate containing a great number of platelets, which is disadvantageous for automated recognition as these may also be misidentified as white cells. Simple centrifugation also results in a concentrate having a large number of platelets and often results in a change in cell distribution i.e., a change in the relative numbers of the different types of white cells in the sample due to trapping of some of the white cells by the red cells as they are centrifuged down. Both leukoadhesion and the use of phagocytic properties also results in a change in the differential distribution of the white cells from that in the original sample. The density gradient techniques require a great deal of time, skill and special equipment to carry out and are not practical in a laboratory processing a large number of samples daily such as is common in large hospitals.

One technique also employed in the prior art to remove platelets from whole blood is known as defibrination. This may be simply defined as the removal of the coagulating or clot producing agents by removal of the platelets from the blood. In that procedure, several glass beads are added to a container of typically five to ten milliliters of fresh whole blood and the tube slowly shaken or rotated for from five to fifteen minutes. The platelets adhere to the glass beads and upon removal of the glass beads from the solution of whole blood, the resulting solution is essentially free of platelets. A leukocyte concentrate can then be obtained from the defibrinated whole blood. This procedure, however, has two major disadvantages. The first is that fresh blood must be used and that the procedure must be carried out immediately after the blood is drawn. If the blood is not defibrinated immediately it will clot and will be useless for a differential count or many other possible laboratory tests. If an anticoagulant is added the procedure Will not Work. The second major disadvantage of defibrination is the loss of leukocytes in the process due to their sticking to the platelet masses. Not only does this reduce the total number of white cells but it can change the white cell differential by preferential removal of one or more types of these cells.

It has also been known in the prior art to utilize dextran dissolved in a dilute solution of NaCl to obtain a leukocyte-rich plasma from a sample of whole blood to which an anticoagulant such as heparin has been added. That method is described in detail in an article by William A. Skooge and William S. Beck entitled Studies on the Fibrinogen, Dextran and Phytohemagglutinin Methods of Isolating Leukocytes, Blood, Volume XI, No. 5, 1956. Although that method does achieve a leukocyte-rich plasma, it also suffers from several disadvantages. First, the platelets are not removed from the leukocyte fraction. Second, although most are removed, a significant number of red cells remain in the final preparation and third, as with most other methods, the proper white cell differential is not maintained because of preferential removal of one or more species of the white cells. The above, last-discussed method does not employ any fixative agent in combination with the dextran. A fixative agent may be defined for the purposes of this specification as a compound which, while compatable with blood in other respects, when added to blood or a cell-containing fraction thereof, immobilizes the cells and facilitates their identification and the performance of certain other tests. The compounds most frequently used are aldehydes such as formaldehyde or glutaraldehyde.

Thus, as can be seen from the above discussion, none of the methods employed in the prior art have achieved a complete blood platelet removal from the other constituents of blood without disturbing the morphology of the other cells or changing the relative numbers of the various types of such cells which remain in the sample.

SUMMARY OF THE INVENTION In accordance with the present invention the problems associated with the removal of the platelet fraction either from whole blood or other constituents of blood without disturbing either the morphology of the remaining cell constituents or the relative numbers of different types of cells in the cell constituents are solved by the use of a unique combination of reagents in a layered separation process. The method of the invention utilizes a sample of whole blood or blood fractions from which the platelets are to be removed. To this amounts of a suitable fixative agent such as formaldehyde (in the preferred embodiment) and dextran are added in a manner which achieves a desired concentration of these reagents in the final solution. After the passage of sufficient time so that the separation reaction may occur, the platelet-containing fraction may be more thoroughly separated from the remainder of the solution by well-known techniques such as mild centrifuging. The desired fraction may then be recovered. Normally, an anticoagulating agent such as EDTA (ethylenediaminetetraacetic acid) or heparin is added to prevent coagulation of the sample before the reagents are added.

In addition to dextran, it has been found that certain concentrations of polyvinylpyrrolidone in combination with the fixative agent may also be employed to achieve comparable results. Also, although formaldehyde is preferred, other fixative agents including glutaraldehyde and benzenaldehyde have been used.

A DESCRIPTION OF THE PREFERRED EMBODIMENT As stated above, the addition of a unique combination of reagents in the separation of blood components has been discovered to allow the complete removal of the platelet fraction without disturbing either the relative numbers of the cellular constituents or the morphology of these cellular constituents. Thus, in the preferred embodiment, it has been discovered that the combination of formaldehyde and dextran provides a unique layered separation in whole blood or fractions thereof containing blood platelets which allows the selective removal of the platelets.

It has also been discovered that the relative concentrations of the dextran and the formaldehyde in the final sedimentation solution in combination with the average or mean molecular weight of the dextran employed are far more important than the particular mode of addition of the dextran and formaldehyde.

It is preferred, however, as with most laboratory techniques wherein a high degree of repeatability is required standard solutions be used such that given specified amounts can be added to a given sample type and size. Thus, the formaldehyde is normally added in the form of formalin which is a 37 percent aqueous solution of formaldehyde and the dextran is normally a 12.5 percent solution in 0.9 percent aqueous sodium chloride. These individual reagents can also be combined into a single additive reagent to be used in the method of the invention.

A further discussion of the effect of concentrations and molecular weight of dextran is given below. The following are specific illustrations of laboratory procedures which may be used to accomplish the method of the present invention.

ILLUSTRATION I One part of a dextran solution of 0.9 percent aqueous sodium chloride is added to four parts of blood containing EDTA anticoagulant agent and sufficient formalin (3 7 percent aqueous solution of formaldehyde) is added to make the final concentration of dextran approximately 1.8 percent and the final concentration of formaldehyde approximately three microliters per milliliter of sample (approximately 0.11 percent) of the total sample. The mixture is then put in a hematocrit tube and spun in a hematocrit centrifuge for from one to five minutes. This results in a distinct layered structure being created in the hematocrit tube which includes a relatively thick layer of red cells on the bottom of the tube, an extremely thin layer of white blood cells adjacent the red cells, a thin layer of platelets above the white cells and the plasma solution on top. The double fraction containing the platelets and white cells is known as the buffy coat. The tube is then broken at the white cell/red cell interface and the buffy coat and a small amount of plasma is allowed to flow out of the top portion of the tube. The platelet layer is formed as a coagulated mass or plug and, if sufficient care is exercised the plug of platelets can be removed intact leaving behind a concentration of white cells and a small amount of plasma for an accurate differential count sample. Tests have indicated that when the white cells and plasma are then mixed, spread on a slide and the differential count taken, this differential count agrees very well with that utilizing whole blood from the same original blood sample.

The final concentrations of both formaldehyde and dextran utilized in the method of illustration I represent optimum concentrations arrived at by a great deal of experimentation. In this experimentation it was found that concentrations of formalin of the order of 1 microliter per microliter of sample (approximately 0.04 percent formaldehyde) resulted in no platelet plug formation at all, while formalin concentrations above about 6 microliters of formalin per milliliter of sample (about 0.24 percent formaldehyde) resulted in undesirable changes in the white blood cell differential count. Thus, while other concentrations within that range will produce a platelet plug without affecting the white cell differential count adversely, approximately 3 microliters per milliliter of sample is found to produce the optimum results.

Extensive experimentation was also conducted in arriving at the final optimum percentage of dextran. Initial experimentation in regard to finding the best dextran concentration included a determination of the range molecular weights of dextran which produce the best sedimentation of red cells and platelets.

Dextrans, of course, are highly branched glucosans normally produced by the fermentation of sucrose solutions by certain bacteria and occur naturally. Dextrans having a wide variety of average or mean molecular weights ranging from a few thousand to several million can be obtained by partial hydrolysis of native dextran.

From the experimentation conducted to date in regard to the present invention, it appears that the range of average molecular weights of dextrans which may be successfully used in the method of the present invention extends from somewhere in the vicinity of 100,000 to somewhat over 1,000,000. Thus, utilizing dextran having an average molecular weight of 20,000 in various percentages in the solution resulted in no sedimentation of red cells or platelets separation and appeared to be well below the lower limit of usable dextran. Dextrans having an average molecular weight of between 250,000 and 500,000, however, produced excellent results in the concentration ranges discussed below. As the average molecular weight of the dextran is increased, however, a decided increase in the viscosity of the solution is noted which may hamper the suitability of high molecular weight dextrans for use in the method of the invention. A dextran having an average molecular weight of approximately two million, is extremely viscous and is extremely difiicult to use in standard volumetric reagent handling apparatus.

Utilizing dextran in the 250,000 to 300,000 average molecular weight range, a study was made to determine the optimum concentration of dextran in the separation of blood platelets. The study was conducted utilizing 16 blood samples in which all 16 blood samples were subjected to sedimentation tests involving a variety of final dextran percentages. In these tests it was observed that a final concentration of 1.2 dextran or below produced little or no platelet separation in the sample. At 1.6 percent dextran little or no separation of platelets or plug formation was observed in two samples, some separation was observed in four samples and a complete separation occurred in ten of the samples. At a concentration of 1.8 percent dextran, complete separation was observed in 14 of the samples and considerable although not complete separation was observed in the two samples which produced no separation at 1.6 percent dextran. At a concentration of 2.0 percent dextran, complete separation and plug formation resulted in the samples, however, three of the samples exhibited some clumping of white cells and in four of the samples quantities of red cells were observed as remaining in the buffy coat. When the concentration of dextran was increased to 2.4 percent, complete platelet separation occurred but high amounts of red cells and clumped white cells were observed in the buffy coat. It was further observed at dextran concentrations of 2.4 percent and above that the white cell layer in the hematocrit tube became split into two portions. The white cells containing an internal granular-like structure or granulocytes were distributed between the red cell and platelet layers and the white cells having no internal granular structure or a granulocytes were located in a separate layer on top of the platelets. That condition, of course, would make a valid differential count impossible. In view of all the experimentation conducted to date, the conclusion has been that a dextran concentration in the range of from about 1.6 to about 1.8 percent presents the optimum dextran concentration.

While the so-called hematocrit method of platelet removal utilized in Illustration I is a possible laboratory technique and was used for the bulk of the optimization work in regard to the method of the invention, a method more suitable both to the individual conducting the tests and to ease of automation has been developed and is illusrated in Illustration II, below.

ILLUSTRATION II Two ml. fresh whole blood containing EDTA anticoagulant was added to a tube containing sufficient formalin and 12.5 percent dextran in 0.9 percent aqueous sodium chloride (average molecular weight 275,000) to provide a final solution having a concentration of approximately 3 microliters formalin per milliter of total solution (approximately 0.11 percent formaldehyde) and 1.8 percent dextran. This mixed solution was then allowed to stand for approximately 15 minutes. The tubes were then batch rotated for approximately 15 minutes to achieve a slight continuing agitation of the mixture. The batch rotation conducted with the tubes in a vertical plane and at very low speed (approximately 10 r.p.m.s). At this point platelets had separated from the plasma and white cells; and the red cells had formed clumps known as rouleaux which were segregated also from the plasma and white cells. The tubes were then subjected to a centrifugation in a low g centrifuge, at approximately 1l0 g for approximately three minutes. When the tubes were thereafter removed from the centrifuge welldefined fractions existed in which the red cells and the platelets were completely separated from the white cells suspended in the plasma. The white cell-plasma layer was located above the red cell and platelet layers at this point and could readily be removed from the tube without disturbing the other constituents. There was no need for breaking the tube as in the hematocrit process and precisely removing a plug of platelets. Instead the plasma solution with suspended white cells could be drawn off without disturbing the other layers.

It further concentration of the leukocytes or white cells is required, the solution of white cells suspended in plasma removed from the low g batch centrifuge may be recentrifuged in a high g (1,0001,500 g) for a period of time, normally about five minutes and the supernatant liquid poured off yielding a concentrated button of White cells. To this concentrated button of white cells a re suspension fluid consisting of 0.9 percent aqueous sodium chloride may be added to achieve the degree of sample dilution desired for the particular test to which the white cell button is to be subjected, usually a differential count. Resuspension can be readily achieved by simply vibrating the tube to break up the white cell button and resuspend the cells in the resuspension fluid. While resuspension is normally accomplished utilizing aqueous sodium chloride other compatable liquids which preserve the sample without afiecting cell morphology may be used. These are all referred to as cell maintenance solutions.

In regard to the method of Illustration II, it might be performed in normal hospital procedures. Prefabricated vacuutubes might even be provided which would contain the mixed formaldehyde and dextran and the blood sample drawn directly from the patient. This, of course, could reduce the laboratory time required in preparing the sample as the initial standing period could start as the sample is drawn.

Some experimentation has been done in regard to Illustration II insofar as variations in concentration of the formaldehyde and dextran are concerned with the result that the optimum concentrations reagents for the multistep method of Illustration H appear to be the same as for the hematocrit method of Illustration I. It was observed, however, that if the concentration of formalin was increased much above 3 microliters per milliter (about 0.11 percent formaldehyde) a decrease was noticed in the relative numbers of neutrophils (the most common type of white cell normally making up about 65 percent of a normal white cell count) was decreased in the final concentrated white cell preparation. This, of course, results in an undesirable condition whereby the final preparation is not representative of the true white cell count involved. At the higher formaldehyde concentration, however, there was a resulting better platelet re moval. The observed effect of varying the dextran concentration appeared to have somewhat the same effect. Thus, an increase in the dextran concentration much above 1.8 percent also resulted in an increasing removal of neutrophils along with the platelets and in some red cell contamination of the final sample. As the concentration was decreased below 1.8 percent the platelet sedimentation aspect was adversely affected but a better white cell retention was obtained.

It was also noticed that when samples were processed eliminating the initial standing time of the mixed solution, this resulted in a marginal overall platelet removal after the ensuing steps were followed. Fifteen minutes appeared to be the minimum acceptable time for consistently good sampling results. If the samples were allowed to stand even longer than fifteen minutes, even more consistent and better platelet removal resulted. The samples could be allowed to stand up to two hours with retention of excellent differential count results. Of course, in any practical laboratory process a two hour stand time is probably longer than most clinical laboratories would desire.

It was also observed that the first centrifuge step had to be an extremely low speed. If this procedure is not followed, the result is a loss of neutrophils from the white cell segment of the segmented sedimentation. For example, a sample which resulted in a good white cell differential count agreement utilizing a centrifugation step at 1l0 g exhibited a loss of 24 percent in the neutrophil count when the same sample was centrifuged at 1,140 g. At the higher centrifugation speed there also appeared to be a loss of total white cells in addition to changes in the relative numbers of the white cells.

It should be noted in regard to both of the outlined Illustrations that wide variation in biological platelet behavior was observed within the sample donor base utilized. Why, for example, behavior for a given concentration of dextran was not completely consistent throughout the entire sample base is unknown. It may well be that there is a range of biological sensitivities to sedimentation reagents such that 100 percent repeatability is not obtainable. It is known that the entire donor base did not consist of blood samples all of which contained normal or near normal white blood cell differential counts. It may well be that along with abnormal white cell differential counts some abnormalities in platelet properties were also present which may have caused the minor number of inconsistent separation results. Generally, the results appear within the range of normal biological reaction diiferences for such tests, however. As a general approach it appears that the method of the present invention represents a great improvement over all the previous procedures heretofore known.

In addition to the above enumerated illustrations samples were run using the same basic formaldehyde concentration but substituting polyvinylpyrrolidone (PVP) having an average molecular weight of approximately 360,- 000 instead of the dextran. A final solution containing 1.2 percent PVP produced an incomplete platelet removal. A higher concentration of 2.4 percent PVP produced good results inasmuch as complete platelet removal was achieved without detrimental effects to the White cell differential count sample. However, there was generally a higher number of residual red cells in the final concentrated white blood cell sample than was observed with the use of dextran.

While no conclusive experimentation has been carried on to date, it is thought highly probable that certain other materials might be useful as platelet sedimentation agents. These materials include fibrinogen, phytohemagglutinin, gelatin, acacia and ficoll. Many of these above enumerated materials are also available in a wide variety vof molecular weight ranges and any optimization would have to be determined by separate extensive testing.

Although the above discussion of the preferred embodiment has been limited to formaldehyde insofar as the fixative agent is concerned, experiments with glutaraldehyde and benzaldehyde indicate that these compounds and probably other compatable fixative agents can be used instead of formaldehyde in the method of the present invention. The Optimum concentration ranges of these other suitable agents in the method of the invention has not been determined to date. Preliminary tests also indicate that these alternative fixative agents though useable may be somewhat inferior to formaldehyde in the method of the invention.

The embodiments of the invention in which an exclusive property or right is claimed are defined as follows: 1. A method for separating the platelet fraction from other constituents of blood, said method comprising the steps of forming a mixture by:

combining with said blood constituents a first additive comprising an amount of a suit able blood fixative agent and a second additive comprising an amount of a compound from a class of compounds which, in combination with said fixative agent, produce a separation of said platelet fraction allowing separation to occur by sedimentation recovering the desired fraction. 2. The method of claim 1 wherein an anticoagulant is combined with said blood constituents.

3. The method of claim 1 wherein said first additive is one selected from a group of fixative agents consisting of formaldehyde, glutaraldehyde and benzaldehyde and wherein said second additive is one selected from the group consisting of dextran and polyvinylpyrrolidone. 4. The method of claim 2 wherein said first additive is formaldehyde and said second additive is dextran.

5. The method of claim 4 wherein said amount of formaldehyde is one sufficient to achieve a final formaldehyde concentration of from about 0.09 to about 0.2 percent and said amount of dextran is one sufficient to achieve a final dextran concentration of from about 1.6 to about 2.6 percent. 6. The method of claim 4 wherein said dextran has an average molecular weight of from about 100,000 to about 1,000,000.

7. The method of claim 4 wherein said dextran has an average molecular weight of from about 200,000 to about 500,000.

8. The method of claim 4 wherein said formaldehyde is in the form of an aqueous solution of formaldehyde.

9. The method of claim 4 wherein said dextran is in the form of a saline solution of dextran in aqueous sodium chloride.

10. The method of claim 4 wherein said dextran added in the form of a solution containing from about 5 percent to about 18 percent dextran in an 0.9 percent aqueous sodium chloride solution.

11. The method of claim 1 further comprising the step of agitating said mixture during said sedimentation step.

12. The method as claimed in claim 1 further including centrifuging said mixture after sedimentation to facilitate recovery of the desired fraction.

13. A method of separating the platelets from other constituents of blood comprising the steps of combining with said blood constituents an amount of formaldehyde sufficient to achieve a final formaldehyde concentration of from about 0.09 to about 0.2 percent an amount of polyvinylpyrrolidone sufficient to achieve a final polyvinylpyrrolidone concentration of from about 2.0 to 3.0 percent allowing sedimentation to occur recovering the desired fraction. 14. The method of claim 13 wherein said polyvinylpyrrolidone has an average molecular weight from about 100,000 to about 500,000.

15. A method of obtaining a platelet-free leukocyte sample from blood comprising the steps of: forming a mixture by combining said blood with an amount of formaldehyde sufiicient to achieve a final formaldehyde concentration of from about 0.09 to about 0.2 percent and an amount of dextran sufficient to achieve a final dextran concentration of from about 1.6 to about 2.6 percent,

allowing said mixture to stand for a minimum of minutes,

agitating said mixture for a period of about 5 minutes or more,

centrifuging said mixture at a force of from about 80 to about 110Xg for from about 2 minutes to about 5 minutes,

recovering the leukocyte-containing fraction from said sedimented centrifuged mixture, centrifuging said leukocyte-containing fraction at a force of from about 900 to about 1500 g separating resulting precipitate from the supernate,

resuspending said precipitate in a cell maintenance solution.

16. The method of claim 15 wherein an anticoagulant is combined with said blood constituents.

17. The method of claim 15 wherein said dextran has an average molecular weight of from about 100,000 to about 1,000,000.

18. The method of claim 15 wherein said dextran has an average molecular weight of from about 200,000 to about 500,000.

19. The method of claim 15 wherein said formaldehyde is in the form of an aqueous solution of formaldehyde.

20. The method of claim 15 wherein said dextran is in the form of a saline solution of dextran in aqueous sodium chloride.

References Cited UNITED STATES PATENTS 2,602,082 5/ 1950 Owen 260209 D 3,212,968 10/ 1965 Fredrick 424- FOREIGN PATENTS 0,716,660 10/ 1954 Great Britain 260209 D OTHER REFERENCES Skoog et al.: Methods of Isolating Leukocytes, May 1956, Blood, vol. 11, No. 5, pp. 436-454.

CHARLES N. HART, Primary Examiner P. A. HRUSKOCI, Assistant Examiner US. Cl. X.R.

2l052, 54; l-l.8; 260209, 209 D