Reduction of Heparin Dose Is Not Beneficial toPlatelet FunctionTakayuki Nakajima, MD, Kohei Kawazoe, MD, Kazuyuki Ishibashi, MD,Yoshimitsu Kubota, CCP, Tatsuya Sasaki, MD, Hiroshi Izumoto, MD, andTohru Nitatori, PhDThird Department of Surgery and Department of Cell Biology and Neuroanatomy, Iwate Medical University School of Medicine,Iwate, Japan

Background. To clarify the effects of the reduction ofheparin dose on platelets, we conducted a prospectivetrial on patients undergoing cardiopulmonary bypass.

Methods. Twenty-three patients undergoing coronaryartery bypass grafting were studied. The systemic hepa-rin dose was 300 IU/kg in the control group (n 5 11) and200 IU/kg in the low-dose group (n 5 12). Heparin-coatedcardiopulmonary bypass equipment was used for boththe groups. Platelet counts, b-thromboglobulin (b-TG)and platelet factor 4 (PF4) concentrations were measuredand the arterial filters in the circuits were observed byelectron microscopy.

Results. Platelet counts were higher in the low-dosegroup than in the control group (p < 0.01). No significant

differences were found in the platelet release reaction(b-TG and PF4). Electron microscopy demonstrated thatcell adhesion on the arterial filters in the control groupwas significantly more marked than in the low-dosegroup (p < 0.01) and that most of the cells on the filterswere neutrophils.

Conclusions. We conclude that the reduction of heparindose with the use of heparin-coated equipment reducesplatelet loss, but does not suppress the platelet releasereaction. Furthermore, the reduction of heparin dosereduces adherence of leukocytes to the filter surface.

(Ann Thorac Surg 2000;70:186–90)© 2000 by The Society of Thoracic Surgeons

Cardiopulmonary bypass (CPB) induces thrombocy-topenia, formation of platelet aggregates, and re-

lease of platelet granule contents, resulting in excessiveblood loss. Platelet dysfunction during CPB is known tobe induced by consumption in the body, blood–gasinteractions, hypothermia, systemic heparinization, andadhesion to the circuits. Therefore, a number of ap-proaches to preserve platelet function have been at-tempted. One of these is a reduction of heparin dose.Previous clinical and experimental studies have reportedthat the reduction of heparin dose with the use ofheparin-coated circuits was reduced platelet loss [1, 2]and blood loss [3, 4]. It is considered that there are twoeffects of the use of heparin-coated equipment and thereduction of heparin dose in this strategy. However, it isnot clear whether the beneficial effects on platelets areinduced by the use of heparin-coated equipment or bythe reduction of heparin dose. We have reported previ-ously that the concomitant use of Duraflo II (Bentley/Baxter, Irvine, CA) heparin-coated equipment with stan-dard systemic heparinization reduces platelet loss andsuppresses the platelet release reaction [5]. To clarify theeffect of a reduced heparin dose on platelets, we com-pared standard-dose heparinization and low-doseheparinization.

Material and Methods

PatientsThis prospective study was performed on 23 consecutivepatients undergoing elective coronary artery bypassgrafting (CABG) using CPB. Patients were stratified ac-cording to the number of distal anastomosis (three orfewer versus more than four). They were then randomlyassigned to the two groups. An assignment was maskedfor a patient, an anesthesiologist and surgeons, but wasnot masked for perfusionists. Eleven patients were as-signed to standard-dose systemic heparinization (controlgroup), and 12 patients to low-dose systemic hepariniza-tion (low-dose group). The following exclusion criteriawere used: (1) preexisting coagulation disorders; (2) on-going antithrombosis therapy with drugs such as aspirinor warfarin within 1 week before the operation; (3) repeatCABG; (4) severe left ventricular dysfunction (ejectionfraction , 0.3); (5) preoperative intraaortic balloon pump;(6) CPB time of less than 90 minutes; and (7) transfusionof more than 3 U of blood during CPB. Informed consentwas obtained from all patients.

Bypass Circuits and OperationIn both groups, all surfaces in potential contact withblood were treated with heparin (Duraflo II). The CPBequipment consisted of a two-stage venous cannula, avenous line, a cardiotomy reservoir (BMR-3500 Gold;Bentley/Baxter), a hollow fiber oxygenator (Univox Gold;Bentley/Baxter), an arterial line, an arterial filter (AF-1040

Accepted for publication Jan 17, 2000.

Address reprint requests to Dr Nakajima, Third Department of Surgery,Iwate Medical University School of Medicine, 19-1 Uchimaru, Morioka,Iwate, 020-8505, Japan; e-mail: t_nakajima@imu.ncvc.go.jp.

© 2000 by The Society of Thoracic Surgeons 0003-4975/00/$20.00Published by Elsevier Science Inc PII S0003-4975(00)01389-8

D; Bentley/Baxter), and a suction tip and line. A cell-saving device was not used for any of the patients.

Five minutes before initiation of CPB, standard-dosesystemic heparinization (300 IU/kg porcine mucosa hep-arin; Novo Nordisk A/S, Denmark) was performed in thecontrol group and low-dose systemic heparinization(200 IU/kg) was used in the low-dose group. Additionalheparin was administered during CPB if the activatedclotting time (ACT) was below 400 seconds in the controlgroup, and below 300 seconds in the low-dose group.After CPB, protamine was administered in the same doseas the initial heparin dose. Prostaglandin E1, whichinhibits surface-induced platelet activation, was not ad-ministered during the operation. Hematocrit was main-tained at above 18% during CPB. Packed red cells wereinfused if the hematocrit dropped to below 18%.

Data CollectionArterial blood samples were taken after induction ofanesthesia, after heparin administration, 5, 20, 60, and 90minutes after the initiation of CPB, at the end of CPB, and10 minutes after protamine administration. The followingparameters were measured: ACT, platelet count, plateletfactor 4 (PF4), and b-thromboglobulin (b-TG) concentra-tions. ACT was measured by Hemochron (Model 401,International Technidyne, Edison, NJ). Platelet countswere corrected for the hematocrit and standardized toprebypass values. PF4 and b-TG concentrations weremeasured by enzyme-linked immunosorbent assay (Di-agnostica Stago, Asnieres-sur-seine, France) as parame-ters of platelet activation.

Sampling For Electron MicroscopySample specimens for scanning electron microscopy(SEM) and transmission electron microscopy (TEM) weredissected from the arterial filter used in the CPB equip-ment after blood perfusion. The samples were rinsedwith 0.9% NaCl solution, and then fixed with 2.5%glutaraldehyde buffered with 0.1 mol/L phosphate buffer(pH 7.2) for 1 hour at 4°C. After rinsing with 7.5% sucrosein the same buffer, the samples were postfixed with 1%OsO4 buffered with 0.1 mol/L phosphate solution (pH7.2) for 1 hour, and dehydrated with serially gradedalcohols. The sample specimens for SEM were furtherimmersed in butylalcohol, and subsequently freezedried. After platinum ion-coating, they were observedunder an electron microscope (S-2300, Hitachi, Japan).The samples for TEM were embedded, and cut intoultrathin sections with an ultramicrotome (Ultracut S,Reichert-Nissei, Japan). After staining with uranyl ace-tate and lead citrate, they were observed under anelectron microscope (H-7100, Hitachi).

As a method of estimating differences in the cellularadhesion on the filter between the two groups, a ten-grade scale was adopted. This was the same method asadopted by Borowiec and associates [6]. It was definedaccording to the morphologic changes of the cells and thedegree of surface coverage by cells and fibrin on SEM:grade 1 denoted no adhesion to the surface and lack ofany morphologic changes. In grade 10 the adhesion was

extremely advanced with large deposits and no distin-guishable cells in a mass.

Statistical AnalysisAll data are reported as the means 6 standard deviations.Differences in preoperative and intraoperative data be-tween the two groups were analyzed using the unpairedStudent’s t test. Comparisons between groups, of cor-rected platelet counts, and b-TG and PF4 concentrations,were made using the repeated-measures analysis ofvariance (ANOVA). Differences in the cellular adhesionas determined by SEM were analyzed by the Mann-Whitney U test for intergroup comparisons. A p value ofless than 0.05 was considered to indicate a statisticallysignificant difference.

Results

Preoperative patient profile and intraoperative details ofboth groups are shown in Table 1. No significant differ-ences in any of the parameters were observed betweenthe two groups. The 24-hour postoperative blood loss,and the number of autologous blood transfusions duringhospitalization are shown in Table 2, with no significantdifferences noted between the two groups. The postop-erative course was uneventful in all patients.

Additional heparin during CPB was not administeredto any of the patients. The mean level of highest ACTduring CPB was 770 6 142 seconds in the control groupversus 490 6 106 seconds in the low-dose group (Fig 1).After administration of protamine sulfate, the ACTsreturned to prebypass levels in both groups.

The platelet counts were significantly lower (p , 0.01)during CPB in the control group whereas in the low-dosegroup, they remained near the base line values (Fig 2).After protamine administration, the platelet counts in thecontrol group were 79% 6 16% compared with 96% 613% in the low-dose group.

The b-TG concentrations increased gradually duringCPB (Fig 3). At the end of CPB, the b-TG concentrationsin the control group were 219 6 56 ng/mL compared with207 6 86 ng/mL in the low-dose group. These differencesbetween the two groups were not significant. Similarly,the PF4 concentrations during CPB increased gradually

Table 1. Preoperative and Intraoperative Profile

ControlGroup

(n 5 11)

Low-DoseGroup

(n 5 12)

Age (y) 60.5 6 4.9 57.8 6 14.8 NSBSA (m2) 1.67 6 0.12 1.74 6 0.12 NSCross-clamp time (min) 82 6 16 81 6 23 NSLowest rectal temperature

(°C)29.3 6 0.8 29.6 6 0.7 NS

Distal anastomoses 3.6 6 1.1 3.4 6 1.0 NS

Mean 6 SD.

BSA 5 body surface area; CPB 5 cardiopulmonary bypass; NS 5not significant.

187Ann Thorac Surg NAKAJIMA ET AL2000;70:186–90 REDUCTION OF HEPARIN DOSE

in both groups (Fig 4). At the end of CPB, the PF4concentrations in the control group were 119 6 32 ng/mLcompared with 122 6 56 ng/mL in the low-dose group.There were no significant differences between twogroups.

After blood perfusion, no large clot formation wasobserved macroscopically on the surface of the arterialfilters in the CPB equipment. The SEM findings of thearterial filters obtained from the low-dose group revealedno cells adherent or a few cells adherent on their surfaces(Fig 5), whereas several cellular and fibrous componentswere found adhering to the surfaces of half of the filtermesh obtained from the control group (Fig 6). The TEMfindings revealed that the morphologic characteristics ofalmost all of these cellular components resembled thoseof neutrophil leukocytes, including the findings of sever-al-segmented nuclei, many specific granules (300 to400 nm in diameter), and azurophil granules (600 to

700 nm in diameter) in their cytoplasm (Fig 7). Thearterial filters obtained from the low-dose group showedless adhesion than the filters obtained from the controlgroup (Fig 8: p , 0.01).

Comment

Systemic heparinization during CPB suppresses activa-tion of the coagulation system and decreases the risk ofcatastrophic intravascular clotting. However, high-doseheparinization is associated with somewhat greater post-operative blood loss [7]. Heparin is known to induceplatelet aggregation [8, 9] and to inhibit platelet function[10]. Chong and associates [11] reported that the degreeof aggregation of platelets increased in a dose-dependentmanner with increasing concentrations of heparin. In thepresent study, platelet counts during CPB in the control

Fig 1. In all patients, additional heparin was not administerd duringcardiopulmonary bypass (CPB). The activated clotting times (ACTs)were significantly higher in the control group than in the low-dosegroup (p , 0.0001; ANOVA).

Fig 2. The platelet (Plt) counts were significantly lower in the con-trol group than in the low-dose group (p , 0.01; ANOVA). (CPB 5cardiopulmonary bypass.)

Fig 3. The b-thromboglobulin (b-TG) concentrations during cardio-pulmonary bypass (CPB) increased gradually in both groups. Therewere no significant differences between the two groups.

Table 2. Blood Transfusions and Blood Loss

ControlGroup

(n 5 11)

Low-DoseGroup

(n 5 12)

Blood usage amountduring CPB

No blood 10 patients(91%)

11 patients(92%)

2 units 1 patient(9%)

1 patient(8%)

During operation(units)

2.7 6 2.4 2.9 6 2.0 NS

Postoperation (units) 2.4 6 1.9 1.3 6 1.3 NSTotal blood transfusion

(unit)5.3 6 3.3 4.4 6 3.2 NS

Postoperative bloodloss at 12 hours (mL)

417 6 181 367 6 188 NS

Mean 6 SD.

CPB 5 cardiopulmonary bypass; NS 5 not significant; 1 unit 5200 mL.

188 NAKAJIMA ET AL Ann Thorac SurgREDUCTION OF HEPARIN DOSE 2000;70:186–90

group were significantly lower than those in the low-dosegroup. It was suggested that the reduction of heparindose suppressed the aggregation of platelets.

After protamine administration, platelet counts in bothgroups decreased. It is reported that infusion of prota-mine sulfate for neutralization of heparin induces throm-bocytopenia [8, 12]. Heparin–protamine complexes acti-vate the classic complement pathway, which causesleukoactivation [13]. The decrease of platelet counts afterprotamine administration indicated that the heparin–protamine interaction induced platelet aggregation.Thus, it was considered that reduction of both heparindose and protamine dose contributed to the reducedextent of platelet loss.

On the other hand, there were no significant differ-ences in the platelet granule release reaction between thetwo groups. This uncoupling of platelet count and plate-let granule release reaction has been reported before [1,2]. Musial and associates [14] reported that a cysteine-rich peptide, which they called “disintegrin,” did notprevent the platelet release reaction, but prevented plate-let adhesion and aggregation. These phenomena sug-

gested that platelet adhesion and the platelet releasereaction occurred by different mechanisms. If the reduc-tion of heparin dose had the same action as “disintegrin,”this uncoupling of platelet count and degranulationcould be explained.

There are few reports on electron microscopic studiesof the effects of reduction of heparin dose along with theuse of heparin-coated circuits [1, 6]. Borowiec and asso-ciates [6] reported that heparin coating of arterial filtersdiminished cellular adhesion to the filter surface duringCPB. Furthermore, they reported that adherence of bloodcells to arterial filters was lower in patients receiving a50% dose of intravenous heparin than in patients receiv-ing a 25% dose. However, they did not examine the cellsadherent on the filters. According to the TEM findings, inthe present study, most adherent cells were neutrophils.Korn and associates [2] reported that in an in vitro modelof a heparin-coated circuit, release of elastase as a markerof neutrophil activation was lower following low-dose

Fig 4. The platelet facter 4 (PF4) concentrations during cardiopul-monary bypass (CPB) increased gradually in both groups. Therewere no significant differences between the two groups.

Fig 5. Lower magnified scanning electron micrograph of the arterialfilter obtained from the low-dose group. (3350 before 5% reduction.)

Fig 6. Lower magnified scanning electron micrograph of the arterialfilter obtained from the control group. (3350 before 5% reduction.)

Fig 7. Highly magnified scanning electron micrograph of the neutro-phil leukocyte on the surface of the arterial filter meshwork (Fm)obtained from the control group. Segmented nucleus (Sn), small spe-cific granules (arrowheads), and azurophil granules (arrows) areobserved in the cytoplasm. (37,500.)

189Ann Thorac Surg NAKAJIMA ET AL2000;70:186–90 REDUCTION OF HEPARIN DOSE

heparinization than following standard-dose hepariniza-tion. Berliner and associates [15] suggested that heparinand C5a might have an effect on the aggregation ofpolymorphonuclear leukocytes. In the present study,concentrations of plasma elastase were not measured.However, the SEM findings suggested that the reductionof heparin dose suppressed activation of leukocytes. Themechanism for the reduced adherence of neutrophils inthe low-dose group remains unknown.

The use of a reduced dose of heparin and heparin-coated circuits was reported to suppress complementactivation [2, 16] and leukocyte activation [16]. Neverthe-less, there are contrary opinions on the practice of reduc-ing the dose of heparin for systemic heparinizationbecause of the increased risk of thrombin generation [1,17]. Bannan and associates [1] reported that anticoagula-tion with a reduced heparin dose (one third of thestandard dose) with heparin-coated circuits was notadequate at 360 minutes. Prolonged perfusion of low-dose heparin may induce consumptive coagulopathy. Inthe present study, as the CPB time in all patients was lessthan 200 minutes (mean, 139 minutes) and the heparindose used was two thirds of the standard heparin dose(200 IU/kg), anticoagulation was probably adequate. Wehave previously reported that the use of heparin-coatedequipment lessened platelet loss and suppressed theplatelet release reaction [5]. However, an additional effectof the reduction of heparin dose on the platelet releasereaction was not recognized in the present study. It isdebatable whether the beneficial effects of the reductionof heparin dose on platelet function outweigh the risk ofthrombin generation.

In summary, the use of a reduced heparin dose andheparin-coated equipment reduced platelet loss, but didnot suppress the platelet release reaction. Furthermore,the reduction of heparin dose reduced adherence of

leukocyte to the filter surface. Further studies are re-quired to clarify the mechanism of adherence ofleukocytes.

References

1. Bannan S, Danby A, Cowan D, Ashraf S, Martin PG. Lowheparinization with heparin-bonded bypass circuits: is it asafe strategy? Ann Thorac Surg 1997;63:663–8.

2. Korn RL, Fisher CA, Livingston ER, et al. Cardiopulmonarybypass, myocardial management, and support techniques.The effects of carmeda bioactive surface on human bloodcomponents during simulated extracorporeal circulation.J Thorac Cardiovasc Surg 1996;111:1073–84.

3. Von Segesser LK, Weiss BM, Pasic M, Garcia E, Turina MI.Risk and benefit of low systemic heparinization during openheart operations. Ann Thorac Surg 1994;58:391–8.

4. Borowiec J, Thelin S, Bagge L, Hultman J, Hansson HE.Decreased blood loss after cardiopulmonary bypass usingheparin-coated circuit and reduction of heparin dose. ScandJ Thorac Cardiovasc Surg 1992;26:177–85.

5. Nakajima T, Osawa S, Ogawa M, et al. Clinical study ofplatelet function and coagulation/fibrinolysis with Duraflo IIheparin coated cardiopulmonary bypass equipment. ASAIOJ 1996;42:301–5.

6. Borowiec JW, Bylock A, van der Linden J, Thelin S. Heparincoating reduces blood cell adhesion to arterial filters duringcoronary bypass: a clinical study. Ann Thorac Surg 1993;55:1540–5.

7. Gravlee GP, Haddon WS, Rothberger HK, et al. Heparindosing and monitoring for cardiopulmonary bypass. J Tho-rac Cardiovasc Surg 1990;99:518–27.

8. Eika C. On the mechanism of platelet aggregation inducedby heparin, protamine and polybrene. Scand J Haematol1972;9:248–57.

9. Salzman EW, Rosenberg RD, Smith MH, Lindon JN, FavreauL. Effect of heparin and heparin fractions on platelet aggre-gation. J Clin Invest 1980;65:64–73.

10. John LCH, Rees GM, Kovacs IB. Inhibition of platelet func-tion by heparin. An etiologic factor in postbypass hemor-rhage. J Thorac Cardiovasc Surg 1993;105:816–22.

11. Chong BH, Ismail F. The mechanism of heparin-inducedplatelet aggregation. Eur J Haematol 1989;43:245–51.

12. Mendeloff EN, Liang IYS, Swain JA, Clark RE. ThromboxaneA2 receptor-specific antagonism in hypothermic cardiopul-monary bypass. Ann Thorac Surg 1994;57:999–1006.

13. Cook JJ, Niewiarowski S, Yan Z, et al. Platelet factor 4efficiently reverses heparin anticoagulation in the rat with-out adverse effects of heparin-protamine complexes. Circu-lation 1992;85:1102–9.

14. Musial J, Niewiarowski S, Rucinski B, et al. Inhibition ofplatelet adhesion to surfaces of extracorporeal circuits bydisintegrins. RGD-containing peptides from viper venoms.Circulation 1990;82:261–73.

15. Berliner S, Fishelson Z, Wasserman L, Pinkhas J, Aronson M.Synergism between zymosan-activated serum and heparinin the induction of polymorphonuclear leukocyte aggrega-tion. Biomed Pharmacother 1988;42:69–72.

16. Ovrum E, Mollnes TE, Fosse E, et al. Complement andgranulocyte activation in two different types of the heparin-ized extracorporeal circuits. J Thorac Cardiovasc Surg 1995;110:1623–32.

17. Edmunds LH, Jr. Surface-bound heparin. Panacea or peril?Ann Thorac Surg 1994;58:285–6.

Fig 8. Cellular adhesion on the arterial filters approximated byscanning electron microscopy views. There were significant differ-ences between two groups (p , 0.01: Mann-Whitney U test).

190 NAKAJIMA ET AL Ann Thorac SurgREDUCTION OF HEPARIN DOSE 2000;70:186–90