It is well recognized that the formed elements of blood are affected by the unphysiologic shear stresses imposed on them by blood handling devices such as pumps, oxygenators, dialysis cartridges, and replacement valves. Erythrocytes (red cells) are most extensively studied in this regard. It is recognized that there is a critical shear stress that will open the cell membrane to release hemoglobin and that this critical stress is a function of exposure time. 1–2 Critical stresses range in magnitude from 103 to 105 to 106, depending on how short the exposure times are. The most common measure of red cell damage by shear stress is the index of hemolysis, the grams of hemoglobin released from the cell’s interior into the blood plasma per 100 L pumped. 3–4 It is recognized that red cells can experience sublethal damage, as well as destruction, from shear stress. 5–6 It has been stated that hemolysis falls into two regimes, one at lower stress where damage is primarily surface mediated, and one at higher stress where bulk flow effects are significant. 7 Additional studies have investigated factors such as surface types, sex and species differences, and blood composition on hemolysis. 8–10 Platelets have received less extensive studies of their shear stress sensitivity as compared with red cells, but they show similar stress and time dependence effects. 11–13 The critical shear stresses are uniformly much lower than those defined for red cells.
Leukocytes (white cells) are the least studied of blood cells with respect to shear stress sensitivity. Some tests have been performed using either white cell suspensions or whole blood. 14–19 Measures of white cell damage have included white cell count, cell membrane integrity, cell adhesiveness, granule staining, or phagocytosis. Exposure times to stress have been long, on the order of 2 to 10 minutes, which is very long-term exposure relative to transit times through most devices. Using a concentric cylinder viscometer to shear whole blood for 10 minutes, Dewitz et al.19 showed significant impact on white cell viability, count, and enzyme release in the range of 100 to 600 dynes/cm2, a much lower value than would be critical for red cells. This is consistent with Takami’s 18in vitro pump test, which showed more impact on white cell viability than on red cell viability.
White cells are critical elements of the immune system. Clinical experience with ventricular assist device implants has shown very high infection rates and declines in B cells and T cells. 20–23 Animal implants show contradictory results, with one report showing sustained suppression of polymorphonuclear (PMN) cell phagocytic function, whereas another study did not show this result. 24,25 These results could be attributed to many factors, among them white cell disruption by unphysiologic shear stresses. A major limitation of existing white cell shear stress sensitivity studies is that the exposure times are extremely long in comparison with what a white cell would experience in an implanted blood pump. Therefore, we conducted a pilot study shearing whole blood at short exposure times, over a range of stresses, and documented the white blood cell count, structural integrity, and phagocytic ability.
Materials and Methods
We chose to apply stress to the blood in a capillary tube test configuration, a classic, easily reproduced fluid flow case. Issues with respect to acceleration and deceleration of a disk are thus avoided. Figure 1 illustrates the test apparatus.
A Harvard 22 Syringe Pump (Harvard Apparatus, Inc., Holliston, MA) was used to propel blood from one syringe to another. In the first test series exploring shear stress response, tubes made with 316 series corrosion resisting steel were used. Tube diameters and lengths were selected, as documented in Table 1, to produce the specified residence times when the pump was operated with a 60 ml syringe at its maximum flow rate of 26.6 ml/min. Table 2 documents characteristics of the flow, as calculated by Fung. 26 The average shear stress on the cells is two thirds of that at the wall.
A viscosity for blood of 0.036 dynes-sec/cm2 and a density of 1.059 gm/cm3 were assumed when required for the calculations. Flows are well within the laminar range, and entrance and exit effects were minimal with the chosen tube lengths. At the shear rates imposed, blood behaves in a Newtonian fashion.
In the second series of tests, tubes 0.069 cm in diameter, 19.8 cm long, and made of polyethylene, Tygon, silicone, or 316 steel were used. The goal was to explore material effects. The tube length and diameter and 16.7 cm3/minute flow resulted in a residence time of 263 ms, about twice that of the longer duration of the first test series, and a wall shear stress of about 312 dynes/cm2, the midrange of Dewitz’s data. In addition to the assays listed below, confocal microscopy was used to image the blood contacting surfaces of the polymeric tubes.
All blood for the first testing series was drawn from the same calf (CCF# 00C-015) in 30 to 60 ml increments. A second calf provided blood for the second series. Testing began approximately 5 minutes after the draw, with some blood reserved as a control. Samples were anticoagulated with 1.5% ethylene diamine tetraacetate (EDTA) in a 9:1 blood to EDTA ratio.
Assays of white cell characteristics immediately followed the shearing.
PMN cells were separated from whole blood immediately after shearing according to Cleveland Clinic Foundation Department of Biomedical Engineering Cell Function Laboratory protocol derived from Carlson and Kaneko. 27
White Blood Cell Count
Cells were stained with trypan blue dye and counted on a hemocytometer.
To determine the number of white cells that were structurally viable postshear, a trypan blue stain was used. An intact leukocyte membrane will exclude this dye, whereas a compromised membrane will not. The protocol for this procedure was based on the work of Coligan et al.28
A fundamental function of a white cell is the ability to phagocytose a foreign body. The ability of the white cells to perform this task was studied with 2.0 μm fluorescent latex particles opsonized in bovine serum solution. The protocol was derived from Oda and Maeda. 29 A functionally viable, or positive, cell is defined as one that ingests more than four particles. This study was conducted using an Olympus model IMT-2 (Tokyo, Japan) inverted phase contrast microscope at ×600 magnification. The phagocytic index is calculated as EQUATION
Table 3 reports the results of a complete blood count performed on the blood donor during the period of testing. The white cell related parameters are slightly above a textbook range, except for lymphocytes (which are a bit low), whereas the red cell parameters are normal except for a high count and a smaller mean corpuscular volume. 30Figure 2 graphically presents the test results. Unstressed control values from all tests constitute a single class of data. These plots show that the white cell count is stable for all test stress levels, whereas the phagocytosis and viability data trend downward. Table 4 lists analysis of variance (ANOVA) results for the listed groups of studies.
Table 5 reports the results of performing t-tests, with a double tailed significance set at α = 0.05. ANOVA analysis across the four different materials tested at the 312 dynes/cm2, 263 ms residence times showed no significant differences for count, viability, or phagocytosis for the materials tested. Confocal microscopy imaging of the polymeric tubes, which were easily opened for viewing, showed no substantial degree of white cell adhesion under these test conditions.
The error bars on the plots overlap to a large degree, and the ANOVA results show little significant difference between the two exposure times. There is a significant difference among the stress levels. The attempt to define breakpoint threshold levels between adjacent levels is problematic. It should be noted that the phagocytosis data appears to show a fairly consistent downward trend except for the last, highest stress level, where phagocytosis improves over the previous value.
Dewitz et al.19 reported significant effects on white cells in the stress range of 100 to 300 dynes/cm2, and nearly complete destruction at 600 dynes/cm2 at very long exposure times. The testing reported here is not strictly comparable with this data because the concentric cylinder device Dewitz and associates used imposes a uniform shear value, whereas the capillary tube shear stresses vary from a maximum at the wall to zero at the centerline. However, issues with the acceleration and deceleration times of the rotor of a concentric cylinder device make short exposure time testing problematic. In terms of practicality, a blood pump would also impose a range of shear stresses on blood cells. Because the goal of this study was to determine the stress values at which significant white cell destruction begins and not to compare the amount of damage, reference to the earlier data still appears useful. Therefore, our study was designed to cover a range of stresses from 100 to 600 dynes/cm2, with an extension to higher values because it was hypothesized that shorter exposure times would increase the tolerable stress. A nominal exposure time of 100 ms was selected as being of the same order of magnitude as our estimates of cell passage time through a rotary dynamic blood pump. The higher and lower exposure times resulted from practical test factors rather than theoretic considerations. The shorter exposure time resulted from a concern with minimizing entrance effects on the laminar flow pattern. For the shortest tube, the estimated entrance length is 15% of the total. 31 The longer exposure time was governed by the ability of the available syringe pump to maintain flow in the longest small-diameter tube. The exposure times differ by about 33%.
In our capillary tube test configuration, the white cells tolerated much higher levels of shear stress than would have produced equivalent deterioration in the 10 minute cone and plate test. White cell count changes were not a significant measure of stress caused damage. In that analysis of structural viability showed a significant change over the stress range for 125 ms exposure, but not for 90 msec exposure, it is tempting to propose that a threshold effect occurs between these exposure times. However, the somewhat anomalous 90 ms data at 133 dynes/cm2 governs this effect. When this time point is removed, the 90 and 125 ms data fit into nearly identical straight lines. The functional test (phagocytic index) shows the strongest response to shear stress, as would be indicated by the slope of a line through the data.
An effort was made to identify differences between the 90 and 125 ms data and to look for dramatic breakpoints between pairs of stresses that might indicate a critical threshold location. For the most part, a difference between the two exposure times could not be demonstrated, at least with our number of samples. The phagocytic index showed a significant difference for the combined data set between the control, zero stress samples, and the 133 dynes/cm2 samples. This possibly represents clearing of a particular subset of white cells. This deterioration of the white cells at physiologic shear stresses may represent an artifact of testing outside the body. The indicated significant difference between 504 and 764 dynes/cm2 is curious, particularly because the mean data for the next highest stress level show a degree of recovery. It must be noted that the 1,240 dynes/cm2 data set is very important for hypothesizing a plateau in this range; without it, a continuously declining line could easily be fit into the phagocytic index–stress level relationship. Over the range of materials, exposure time, and stress tested, the surface of the capillary tube may not be an important factor, as suggested by our inability to show a difference for the materials tested. Relatively minor attachment of white cells to the polymeric tubes was found on the polymeric tube blood contacting surfaces. The lack of difference in test results with the steel tube suggests that that surface could not have accumulated significantly greater numbers of cells. The steel tube tests in the materials comparison study at a third exposure time did show differences in comparison to the stress comparison series. However, the series used a different animal’s blood. Comparing the control values with the sheared blood values of the materials comparison series did not show a significant change between the steel tube sheared results and the unsheared control blood, and therefore a three exposure time comparison was not performed in this study.
This investigation is preliminary, and caution is required with respect to conclusions. Testing with many more samples would be desirable, and one could ask for many comparison and investigations beyond the scope of the current program. However, one can draw some conclusions with which to guide the design of future studies:
- A functional assay may be the more appropriate way to evaluate the effect of shear stress on white cells than a simple cell count, possibly even a dye exclusion structural viability test. In the current testing, the phagocytic index was the more sensitive index of white cell deterioration.
- As with red cells and platelets, white cell stress response should be studied at stress levels and exposure times representative of the clinical situations that might occur with a device. This testing showed the initiation of cell damage at higher stresses, with much shorter exposure times than shown by the Dewitz data. 17
- The phagocytic index appears to decline by large amounts at a stress level and exposure time that would be tolerated by a red cell.
The infection problems with blood pumps are undoubtedly a multifactorial issue, but the possibility that unphysiologic mechanical stresses in some pumps are handicapping the body’s ability to respond to bacterial invasion can not be ignored. More work must be done to characterize the white cell’s response to stress and the effect of blood pumps on white cell function.
This research was funded by a Research Experiences for Undergraduates grant from the National Science Foundation with additional support by the Cleveland Clinic Foundation. Significant assistance with procedures was provided by Carol Culler.
1. Leverett LB, Hellum JD, Alfrey CP, Lynch EC: Red blood cell damage by shear stress. Biophysical J 12: 257–273, 1972.
2. Blackshear PL: Mechanical hemolysis in flowing blood, in Fung YC, Perrone N, Anliker Q (eds), Biomechanics: Its Foundations and Objectives
. Englewood Cliffs, NJ, Prentice Hall, 1972, p. 501.
3. Koller T, Hawrylenko A: Contribution to the in vitro testing of pumps for extracorporeal circulation. J Thorac Cardiovasc Surg 54: 22–9, 1967.
4. Naito K, Mizuguchi K, Nose Y: The need for standardizing the index of hemolysis. Artif Organs 18: 7–10, 1994.
5. Indeglia RA, Shea MA, Forstrom R, et al: Influence of mechanical factors on erythrocyte sublethal damage. Trans ASAIO 14: 264–71, 1968.
6. Sutera SP: Flow-induced trauma to blood cells. Circ Res 41: 2–8, 1977.
7. Leverett LB, Hellums JD, Alfrey CP, et al: Red blood cell damage by shear stress. Biophysical J 12: 257–73, 1972.
8. Bernstein EF, Castaneda AR, Varco RL: Some biologic limitations to prolonged blood pumping. Trans ASAIO 11: 118, 1965.
9. Nichols AR, Williams MC: Suppression of shear-induced hemolysis by three plasma proteins. Biomat Med Dev Artif Organs 4: 21–48, 1976.
10. Indeglia RA, Shea MA, Varco RL, et al: Mechanical and biologic considerations in erythrocyte damage. Surgery 62: 47–55, 1967.
11. Hung TC, Hochmuth RM, Joist JH, et al: Shear induced aggregation and lysis of platelets. Trans ASAIO 22: 285, 1976.
12. Johnson GG, Marzee V, Berstein EF: Effects of surface injury and shear stress on platelet aggregation and serotonin release. Trans ASAIO 21: 413, 1975.
13. Colantuoni G, Hellums JD, Moake JL, Alfrey CP Jr: The response of human platelets to shear stress at short exposure times. Trans ASAIO 23: 626–31, 1977.
14. Dewitz TS, McIntire LV, Martin RR, Sybers HD: Enzyme release and morphological changes in leukocytes induced by mechanical trauma. Blood Cells 5: 499–512, 1979.
15. Kusserow B, Larrow R, Nichols J: Perfusion- and surface-induced injury in leukocytes. Fed Proc 30: 1516–20, 1971.
16. McIntire LV, Dewitz TS, Martin RR: Mechanical trauma effects on leukocytes. Trans ASAIO 22: 444–449, 1976.
17. Salzman EW (ed). Interaction of the Blood with Natural and Artificial Surfaces.
pp. 119–138. New York: Dekker, 1981.
18. Takami Y, Yamane S, Makinouchi K, Glueck J, Nose Y: Mechanical white blood cell damage in rotary blood pumps. Artif Organs 21: 138–42, 1997.
19. Dewitz TS, Hung TC, Russell RM, McIntire LV: Mechanical Trauma in Leukocytes. J Lab Clin Med 90: 728–736, 1977.
20. Gordon SM, Schmitt SK, Jacobs M, et al: Nosocomial bloodstream infections in patients with implantable left ventricular assist devices. Ann Thorac Surg 72: 725–730, 2001.
21. Termuhlen DF, Pennington DG, Roodman ST, et al: T cells in ventricular assist device patients. Circulation 80 (5 Pt 2): III174–III182, 1989.
22. Wellhausen SR, Ward RA, Johnson GS, DeVries WC: Immunological complications of long-term implantation of a total artificial heart. J Clin Immunol 8: 307–318, 1988.
23. Stelzer GT, Ward RA, Wellhausen SR, et al: Alterations in select immunologic parameters following total heart implantation. Artif Organs 11: 52–62, 1987.
24. Paping R, Webster LR, Stanley TH, Razecca K, Kolff WJ: White blood cell phagocytosis after artificial heart implantation. Trans ASAIO 24: 578–80, 1978.
25. Harasaki H, Fukamachi K, Benavides M, et al: A comprehensive hematologic study in calves with total artificial hearts. ASAIO J 41: M266–71, 1995.
26. Fung YC:The Flow Properties of Blood: Biomechanics Mechanical Properties of Living Tissues.
New York: Springer-Verlag, 1993.
27. Carlson GP, Kaneko JJ: Isolation of leukocytes from bovine peripheral blood. Proc Soc Exp Biol Med 142: 853–856, 1997.
28. Coligan JE, Kruisbeek AM, Shevach EM, Strober W (eds):National Institutes of Health. Current Protocols in Immunology. Trypan Blue Exclusion Test of Cell Viability
, Vol. 1, Appendix A3.3. Washington DC: Library of Congress, 1992.
29. Oda T, Maeda H: A new simple flourometric assay for phagocytosis. J Immunol Methods 88: 175–183, 1986.
30. Gross DR:Animal Models in Cardiovascular Research
, 2nd revised ed. Norwell, MA: Kluwer Academic Publishers, 1994, pp. 3–4.
31. Fox RW, McDonald AT:Introduction to Fluid Mechanics
, 4th ed. New York: John Wiley and Sons, 1992, p. 322.