B L O O D C O M P O N E N T S

Sickle Hb polymerization in RBC components from donors withsickle cell trait prevents effective WBC reduction by filtration

David F. Stroncek, Tobie Rainer, Victoria Sharon, Karen M. Byrne, Constance T. Noguchi,

Harvey G. Klein, Alan N. Schechter, and Susan F. Leitman

BACKGROUND: RBC components collected from do-nors with sickle cell trait frequently occlude WBC-reduc-tion filters. In vitro, sickle trait RBCs have the potentialfor sickle Hb (Hb S) polymerization at low oxygen satu-rations and high Hb concentrations.STUDY DESIGN AND METHOD: To determine if thelow pH and high osmolarity of the CP2D used in thecollection contributed to filter failures, the filterability ofsickle trait donor RBCs collected in CP2D was com-pared with RBCs from the same donors collected inheparin.RESULTS: Five of six sickle trait components collectedin CP2D did not complete filtration, but all six RBC com-ponents collected in heparin filtered completely. RBCcomponents collected in CP2D from four other sickletrait donors were divided in two, and one-half wastreated with carbon monoxide to convert Hb S to its li-ganded form to prevent Hb S polymerization. All fourcarbon monoxide-treated components filtered within 9minutes, but only one untreated component filteredcompletely. RBC components collected by apheresiscontained less CP2D, and five of seven sickle traitapheresis components filtered completely; four of thefive filtered rapidly (<15 min) and one filtered in 100minutes. Hb oxygen saturation was greater in the fourrapidly filtering apheresis RBC components (68 ± 9%)than in the three that filtered slowly or incompletely(37 ± 5%, p = 0.03).CONCLUSIONS: Hb S polymerization appears respon-sible for RBC WBC-reduction filter failures. Citrate anti-coagulant and low oxygen saturation are responsible inpart for Hb S polymerization in this setting.

The reduction of WBCs in cellular blood compo-nents may benefit transfusion recipients bypreventing alloimmunization, febrile reactions,cytomegalovirus infections, and possibly trans-

fusion-associated immune suppression.1,2 As a result,many blood centers and transfusion services have begunto provide only WBC-reduced RBC and platelet compo-nents.

WBCs are removed from RBC components by filtersspecially designed for this purpose.3 Although these fil-ters are highly effective, approximately 1 percent of fil-tered RBC units do not meet the FDA criteria for WBCreduction because the quantity of WBCs remaining in theunit or the loss of RBCs is too high.4 Several recent pre-liminary reports indicate that RBC components that donot meet the criteria for WBC reduction are more likely tobe derived from people with sickle cell trait (Hb AS).5-8

One study found that approximately half of the RBC com-ponents collected from people with sickle trait occludeWBC-reduction filters, one-quarter pass completelythrough the filter but the quantity of WBCs remainingexceeds criteria for WBC reduction, and one-quarter aresuccessfully WBC reduced.7

People with sickle cell disease (Hb SS) are homozy-

ABBREVIATIONS: Hb AA = without Hb S; Hb AS = with sickle

cell trait; Hb S = sickle Hb; Hb SS = with sickle cell disease;

HPLC = high-performance liquid chromatography.

From the Department of Transfusion Medicine, Warren G.

Magnuson Clinical Center, and the Laboratory of Chemical

Biology, National Institute of Diabetes and Digestive and

Kidney Diseases, National Institutes of Health, Bethesda,

Maryland.

Address reprint requests to: David F. Stroncek, MD, Na-

tional Institutes of Health, Warren G. Magnuson Clinical Cen-

ter, Department of Transfusion Medicine, Building 10, Room

1C711, 10 Center Drive MSC-1184, Bethesda, MD 20892-1184;

e-mail: dstroncek@dtm.cc.nig.gov.

Received for publication February 28, 2002; revision re-

ceived May 14, 2002, and accepted May 30, 2002.

TRANSFUSION 2002;42:1466-1472.

1466 TRANSFUSION Volume 42, November 2002

gous for Hb S and experience chronic anemia, acutechest syndrome, stroke, pain crises, splenic dysfunction,and renal dysfunction due to the intracellular polymer-ization of Hb S and resulting pathophysiologic pro-cesses.9-11 In contrast, people with Hb AS are heterozy-gous for Hb S and have approximately 40 to 60 percentHb A in each RBC. They experience no sickle cell crisesand few other manifestations of sickle cell disease.9-11 HbS in RBCs from people with Hb AS can polymerize andcause cell sickling at low oxygen tension, very low pH,and/or markedly increased cellular Hb concentration,but under physiologic conditions the concentration ofHb S in RBCs from people with Hb AS is not great enoughto cause polymerization.12-15 One exception is in the hy-perosmotic kidney and renal concentrating defect insickle trait that is dependent on the percentage of Hb S.16

Blood collected by phlebotomy flows into bags con-taining 63 mL of citrate anticoagulant preservative, whichis hyperosmotic and has a low pH. The extreme condi-tions of the citrate anticoagulant may damage the firstportion of the blood collected and result in a so-called“citrate collection lesion.”17 We hypothesized that the in-effective filtration of RBCs from people with Hb AS is duelargely to the collection of blood into citrate anticoagu-lant preservative, which results in the intracellular poly-merization of Hb S. The low oxygen tension that is foundin collected blood or that results from subsequent deoxy-genation during storage and low pH and high osmolarityof the citrate anticoagulant solution could promote Hb Spolymerization. The purpose of this study was to deter-mine the role of Hb S polymerization in citrate-anticoagulated RBC components from Hb AS individualsin the performance of WBC reduction by filtration.

MATERIALS AND METHODSStudy designThe goal of the first part of the study was to determine ifa CP2D “collection lesion,” caused by the low pH (5.7)and high osmolarity (585 mOsm/kg) of the CP2D, con-tributed to filter failures. From each healthy donor stud-ied, 250 mL of blood was collected into 31 mL of CP2D,and 250 mL of blood was collected into 2.5 mL (2500units) of heparin. Both donors with Hb AS and withoutHb S (Hb AA genotype) were studied. RBCs were pre-pared, AS-3 was introduced, and the RBC componentswere filtered. Blood counts, pH, oxygen tension, and os-molarity were assessed before and after filtration. Filtra-tion time, filtration volume, RBC recovery, and residualWBC counts were compared.

To determine the role of Hb S polymerization in filterfailure, 500 mL of blood was collected by phlebotomyinto 63 mL of CP2D. RBCs were prepared and AS-3 wasadded. The component was divided into two parts; onehalf was treated with carbon monoxide gas for 60 min-

utes before filtration to convert Hb S to its liganded con-figuration, which prevents Hb S polymerization, and theother half was filtered without further treatment.

In the second part of the study, RBCs were collectedby apheresis to determine if reduced exposure to CP2Dwould prevent filter failures. Each unit of apheresis RBCswas split in half. One half of the unit was filtered imme-diately, and its properties and filterability were comparedwith those of components collected by phlebotomy. Theother one-half unit of apheresis RBCs was used for otherstudies.

DonorsThese studies were approved by a National Institutes ofHealth Institutional Review Board, and informed consentwas obtained before the blood was collected. Donors metall AABB criteria for donating whole blood. Sickle cell traitwas confirmed by ion exchange high-performance liquidchromatography (HPLC) analysis of donor RBCs (VarianHPLC system [�-thalassemia short program], Bio-Rad Di-agnostics Group, Hercules, CA). All donors were asked todisclose their current smoking status.

Collection and filtration of phlebotomyRBC componentsRBCs were collected into a modified collection bag setthat included CP2D, AS-3, and an RBC WBC-reductionfilter (RCM1, Leukotrap RC System, Medsep Corporation,Pall Medical, Covina, CA). The set was altered by remov-ing half (31 mL) of the CP2D from the collection bag,removing half (50 mL) of the AS-3, adding an additionalcollection bag that contained sodium heparin (2.5 mL,1000 units/mL; Elkins-Sinn, Cherry Hill, NJ), adding a bagcontaining AS-3 (50 mL), and adding a second WBC-reduction filter (RCM1, Pall Medical). From each donor, avolume of 250 mL of whole blood was collected into thebag containing CP2D, and 250 mL was collected into thebag with heparin. The bags were rocked during collection(Sebra, Tucson, AZ).

RBCs were prepared by centrifugation of wholeblood collected in CP2D and heparin, and RBCs werefiltered according to the manufacturer’s instructions,with the exception that only one half of the AS-3 (50 mL)was added to the RBCs. Components were defined asfiltering completely if all RBCs drained out of the upperprefiltration bag and into the final RBC storage bag.Samples were taken before and after the addition of AS-3and after filtration for measurement of complete bloodcounts, osmolarity, blood gases, and pH.

Carbon monoxide treatment of phlebotomyRBC componentsOne unit of blood was collected in CP2D, RBCs wereprepared, and AS-3 was added. The RBC component was

Hb S POLYMERIZATION CAUSES RBC FILTER FAILURES

Volume 42, November 2002 TRANSFUSION 1467

then divided in half. One half was filtered as describedabove (RCM1, Pall Medical), and the other half wastreated with carbon monoxide before filtration. To treatthe component with carbon monoxide, the half unit wasadded to a tonometer (Fisherbrand septum-port gas sam-pling tube, 250 mL, Fisher Scientific, Pittsburgh, PA), andcarbon monoxide (Aldrich, St. Louis, MO) was allowed toflow slowly through the tonometer at room temperaturefor 60 minutes as the tonometer was gently rocked in anexhaust hood. After 1 hour of incubation with carbonmonoxide, the RBCs were transferred from the tonometerto a bag and filtered with a RBC WBC-reduction filter(RCM1). The filter was primed with AS-3, but rather thanadding the AS-3 to the RBCs after the AS-3 passedthrough the filter, the AS-3 was diverted to an empty bagas it left the filter.

Collection and filtration of apheresisRBC componentsRBCs were collected using a blood cell separator (MCS +,Haemonetics, Braintree, MA) according to the manufac-turer’s recommendations, with the exception that only 1unit of RBCs was collected. CP2D was added to wholeblood as the blood was withdrawn from the donor’s vein,at a ratio of 1 part CP2D in 16 parts blood. Immediatelyafter the RBCs were collected, the blood cell separatoradded the RBCs to a bag containing 100 mL of AS-3.

The apheresis RBC component was divided into twoparts. One half was filtered within 2 hours of collectionusing an RBC WBC-reduction filter (RCM1). Before theRBCs were filtered, the filter was primed with AS-3 (asdescribed above). The second half was used for otherstudies.

Blood counts, blood gases, and osmolaritiesBlood counts were measured with an automated cellcounter (Cell-Dyn 4000, Abbott Diagnostics, Santa Clara,CA). Blood gases, pH, sodium, potassium, chloride, bi-carbonate, and glucose levels were measured with ablood gas analyzer (Radiometer ABL 700 Series, Radiom-eter Analytical SA, Lyon, France). Osmolality measure-

ments were performed with a PSI-Multi-Osmette model2430 instrument (Precision Systems, Natick, MA).

Statistical analysisValues represent the mean � 1 SD. Groups were com-pared using the students t-test. In some cases, pairedt-tests were used.

RESULTS

Filtration of phlebotomy RBC componentscollected in CP2DRBC components from six donors with Hb AS were stud-ied. All donors were healthy and met eligibility criteria forallogeneic blood donation. The donor’s mean age was 46years (range, 32-53), one was male, and all were AfricanAmerican. The percentage of Hb S ranged from 33.7 to39.0 percent. One Hb AS donor smoked cigarettes.

CP2D RBC components collected from five of the sixHb AS donors occluded the filter before all the RBCspassed through (Table 1). In two of these five donors, flowthrough the filter was completely blocked before anyRBCs could pass into the collection bag. The Hb AS donorwhose CP2D RBC component filtered completely was theonly one who smoked cigarettes, and filtration time was72 minutes. The RBC recovery of this donor’s componentwas 71 percent, and the residual WBC count was 0.11 �

106 cells.RBC components were collected from six Hb AA

healthy African Americans; their mean age was 39 years(range 32-49) and three were male. All six control Hb AARBC components collected in CP2D filtered completely(data not shown). The mean filtration time was 18 � 5minutes and ranged from 11 to 26 minutes. The residualWBC counts in all six components were less than 0.15 �

106 cells. The RBC recovery of the CP2D components col-lected from the Hb AA donors was greater than the RBCrecovery of components collected from the Hb AS donors(82 � 4% vs. 26 � 27%, respectively; p < 0.004).

TABLE 1. Filtration of Hb AS RBC components collected in CP2D and heparin

ComponentHb S(%)

CP2D Heparin

Filtrationoutcome

Initialvolume

(mL)

Filtrationtime(min)

RBCrecovery

(%)

ResidualWBCs(×106)

Filtrationoutcome

Initialvolume

(mL)

Filtrationtime(min)

RBCrecovery

(%)

ResidualWBCs(×106)

1 35.7 obstructed 155 >120 34 NA complete 131 20 96 0.062 39.0 obstructed 120 >120 0 NA complete 123 54 75 0.043 35.4 complete 147 72 71 0.11 complete 152 28 68 0.54 38.8 obstructed 113 >120 26 NA complete 116 25 69 0.135 34.3 obstructed 156 >120 0 NA complete 137 16 80 0.316 33.7 obstructed 129 >120 40 NA complete 112 12 83 0.11

Controls (n = 6) 0 complete 110-151 18 ± 5 82 ± 4 <0.15 complete 70-135 11 ± 5 91 ± 10 <0.66

STRONCEK ET AL.

1468 TRANSFUSION Volume 42, November 2002

Filtration of phlebotomy RBCcomponents collected in heparinTo determine if CP2D contributed to fil-ter occlusion in donors with Hb AS,one-half unit of blood was collectedinto heparin by phlebotomy from thesame six donors with Hb AS. All six HbAS RBC components collected in hepa-rin filtered completely (Table 1). The re-sidual WBC count in all six componentswas less than 1 � 106 cells. The RBCrecoveries of the six components col-lected in heparin were greater than theRBC recoveries of the six componentscollected in CP2D from the same do-nors (78 � 10% vs. 26 � 27%, p <0.005).

RBC components from the six do-nors with Hb AA (normal African Americans) were alsocollected in heparin and filtered. When the RBC recover-ies and filtration times in the Hb AA donors were com-pared with those of the six components collected in hep-arin from Hb AS donors, no differences were found infiltration time (26 � 15 min vs. 11 � 5 min, p = 0.07) orRBC recoveries (78 � 10% vs. 91 � 10%, p = 0.10).

Comparison of the properties of RBC componentscollected in CP2D and heparinThere was no difference in pH, osmolarity, oxygen ten-sion, Hb oxygen saturation, or MCHC between the Hb AScomponents collected in CP2D and heparin (Table 2).The similarities of these parameters among the RBC com-ponents collected in CP2D and heparin suggest that acitrate collection lesion was contributing to the filter fail-ures.

Blood chemistry levels, pH, osmolarities, and oxygensaturations measured in whole blood were comparedamong components collected from Hb AS donors andfrom Hb AA donors (Table 2). When components werecollected in CP2D, sodium, glucose, and osmolarity werelower and chloride, potassium, and pH were higher incomponents collected from Hb AS donors than in controlHb AA donors. There was no difference between the twogroups in sodium, chloride, glucose, osmolarity, and pHamong components collected in heparin. There was nodifference in oxygen tension or oxygen saturation amongHb AS and Hb AA components collected in CP2D andheparin. The differences in laboratory parameters amongHb AS and Hb AA components collected in CP2D, but notthose collected in heparin, suggest that CP2D had agreater effect on RBCs from donors with Hb AS than withHb AA donors.

When blood is collected into CPD, the RBCs swell.16

As expected, when RBC indices were compared amongRBC components collected from Hb AA donors in CP2D

and from Hb AA donors in heparin, MCV was greater, andMCHC was less in CP2D components (Table 2). In con-trast, no increase in RBC volume occurred when Hb ASRBCs were collected into CP2D. There was no differencein Hb AS donor MCV or MCHC between RBC compo-nents collected in CP2D and those collected in heparin.

Filtration of carbon monoxide-treatedRBC componentsTo determine if Hb S polymerization was responsible forthe occlusion of WBC-reduction filters, RBC componentswere treated with carbon monoxide to convert Hb S to itsliganded form, which prevents polymerization. RBC com-ponents collected in CP2D from four Hb AS donors weredivided in half; one half was treated with carbon monox-ide before filtering and the other half was filtered withoutfurther treatment (Table 3). All four RBC componentstreated with carbon monoxide filtered completely, butonly one of the four untreated components filtered com-pletely. In addition, two of the six CP2D Hb AS RBC com-ponents collected in the first part of these studies did notfilter at all. One component was hemolyzed and was nottested further, but the other component was treated withcarbon monoxide and filtered again. After carbon mon-oxide treatment, this component filtered completely in 7minutes, with a RBC recovery of 85 percent and a residualWBC count of 0.04 � 106 cells. The RBC recoveries of thefive carbon monoxide-treated Hb AS components weresignificantly greater than those of the untreated compo-nents from the same donors (84 � 4% vs. 32 � 36%, p <0.04).

Filtration of apheresis RBC componentsApheresis RBC components were collected from sevendonors with Hb AS. All seven donors were African Ameri-can, their median age was 39 years (range, 20-50), and

TABLE 2. Comparison of chemistry levels* in whole-blood componentscollected in CP2D and heparin in Hb AS and Hb AA donors

Number

CP2D Heparin

Hb AS6

Hb AA6

Hb AS6

Hb AA6

Sodium (mM ) 142 ± 9 152 ± 4† 138 ± 2 139 ± 2Potassium (mM ) 4.5 ± 0.9 3.2 ± 0.2† 4.8 ± 0.7 4.1 ± 0.2Chloride (mM ) 97 ± 9 80 ± 3† 105 ± 1 105 ± 1Glucose (mg/dL) 277 ± 197 603 ± 88† 100 ± 21 100 ± 14Osmolarity (mOsm/kg) 300 ± 20 335 ± 7† 291 ± 4 297 ± 5Oxygen saturation (%) 58.2 ± 23.4 53.7 ± 17.2 50.1 ± 16.1 43.5 ± 21.2pH 7.22 ± 0.14 6.97 ± 0.08† 7.27 ± 0.12 7.29 ± 0.04MCV (fL) 84.3 ± 3.6 91.2 ± 6.7‡ 83.9 ± 3.1 87.4 ± 5.1MCH (pg) 28.7 ± 1.2 28.8 ± 2.0 28.4 ± 1.4 28.9 ± 1.6MCHC (g/dL) 33.4 ± 9 31.6 ± 1.5‡ 33.8 ± 0.6 33.0 ± 0.9

* Chemistries and RBC indices were measured in components at the time of collection.† p < 0.02 for comparison of CP2D blood collected from Hb AS and Hb AA donors.‡ p < 0.02 for comparison of CP2D and heparin components collected from Hb AA

donors.

Hb S POLYMERIZATION CAUSES RBC FILTER FAILURES

Volume 42, November 2002 TRANSFUSION 1469

two were men. Five of the apheresis RBC componentspassed completely through the WBC-reduction filter.Four of these five apheresis RBC components filtered inless than 15 minutes and one filtered in 100 minutes. Theresidual WBC count in all five components filtering com-pletely was less than 0.9 � 106 cells, and RBC recovery inall five was 85 percent or greater (Table 4). When the RBCrecovery of components collected from Hb AS donors byapheresis was compared with components collectedfrom Hb AS donors by phlebotomy into CP2D, RBC re-covery was greater in the apheresis RBC components(67� 35% vs. 28 � 27%, p < 0.05).

As a control, apheresis RBC components were col-lected from four healthy Hb AA male donors; three wereAfrican Americans and one was Caucasian. Their meanage was 38 years (range, 28-46). For all four controlapheresis RBC components, filtration time was less than10 minutes, RBC recovery was greater than 87 percent,and residual WBC counts were less than 0.3 � 106 cells.RBC recoveries were similar for apheresis RBC compo-nents collected from Hb AS and Hb AA donors (67 � 35%vs. 91 � 32%, p = 0.22).

Properties of RBC components collectedby apheresisAlthough several differences were found in the propertiesof CP2D components collected by phlebotomy amongHb AS and control donors, there were no differences inchemistries, RBC indices, pH, osmolarity, or oxygen satu-ration (sodium, potassium, chloride, glucose, MCV,

MCH, and MCHC) among apheresisunits collected from Hb AS and Hb AAdonors (data not shown).

Comparison of apheresiscomponents filtering rapidly withthose filtering slowlyThe two Hb AS apheresis componentsthat occluded the filter and the one thatcompleted filtration in 100 minuteswere considered “poorly” or “slowly”filtering components. The properties ofthe four apheresis Hb AS componentsthat filtered “rapidly” (in less than 15

minutes) were compared with those of the three unitsthat filtered “poorly.” There was no difference in Hb Sfraction, pH, or MCHC between the two groups, but theoxygen saturation was greater in the rapidly filteringgroup (68 � 9% vs. 37 � 5%, p < 0.03) (Fig. 1).

DISCUSSION

Polymerization of Hb S during the collection and pro-cessing of blood appears to be responsible for the inef-fective performance of RBC WBC-reduction filters withRBC components collected from Hb AS donors. Becauseoxygenated or liganded Hb S does not polymerize, thefinding that Hb AS components treated with carbonmonoxide can successfully undergo WBC reduction byfiltration suggests that Hb S polymerization is responsiblefor filter performance problems. When Hb S polymerizes,RBC intracellular viscosity increases, reducing deform-ability and impairing filterability.18 The trapping of RBCswith polymerized Hb S in WBC-reduction filters leads toeither complete obstruction of flow or channeling of flowthat makes filtration ineffective.

These findings are significant because the Hb S po-lymerization in RBCs from persons with Hb AS was notthought to be of clinical relevance during blood donation.In healthy donors, oxygen saturation is generally consid-ered high enough to prevent Hb S polymerization. How-ever, this study shows that nonphysiologic conditionspresent in blood collection bags can be associated withHb S polymerization.

TABLE 3. Filtration of Hb AS RBC components collected in CP2D and treated with carbon monoxide

Component

Untreated components Carbon monoxide-treated components

Filtrationoutcome

Filtrationtime(min)

RBCrecovery

(%)

ResidualWBCs(×106)

Filtrationoutcome

Filtrationtime(min)

RBCrecovery

(%)

ResidualWBCs(×106)

7 complete 26 79 0.715 complete 4 84 0.018 obstructed >120 0 NA complete 7 88 0.019 obstructed >120 19 NA complete 9 76 0.26

10 obstructed >120 62 NA complete 9 86 0.01

TABLE 4. Filtration of RBC components collected by apheresis fromHb AS donors

ComponentHb S(%)

Filtrationoutcome

Initialvolume

(mL)

Filtrationtime(min)

RBCrecovery

(%)

ResidualWBCs(×106)

1 39.0 complete 162 12 86 0.412 37.3 complete 162 8 85 0.043 30.4 complete 160 10 91 04 39.2 obstructed 155 >120 13 NA5 38.3 complete 158 6 93 0.896 39.2 obstructed 160 >120 17 NA7 35.1 complete 160 100 85 0.24

Controls (n = 4) 0 complete 155-169 8 ± 2 91 ± 3 <0.3

STRONCEK ET AL.

1470 TRANSFUSION Volume 42, November 2002

The situation is perhaps closest to what happens inthe renal medulla, where oxygen tension and pH are lowand osmolarity is high. The high osmolarity draws intra-cellular water from RBCs, increasing the Hb S concentra-tion. The increase in Hb S concentration and the lowoxygen tension can result in Hb S polymerization, micro-vascular occlusion, and impaired ability to concentrateurine.16 In an analogous manner, collection of blood intoa hyperosmotic environment can expose RBCs to condi-tions that result in polymerization of Hb S.

Although several factors affect the polymerization ofHb S, the major cause of Hb S polymerization duringwhole-blood phlebotomy is likely the collection into ci-trate anticoagulant solution. Although all six RBC com-ponents collected into heparin were effectively WBC re-duced by filtration, only one of six units collected intocitrate anticoagulant was effectively WBC reduced. Whenwhole blood is collected by phlebotomy, it is collectedinto 63 mL of CP2D or a similar citrate-based anticoagu-lant. CP2D has an osmolarity of 585 mOsm per kg and apH of 5.7. These properties of CP2D and the low oxygensaturation of venous blood favor Hb S polymeriza-tion.12,13 The effects of citrate anticoagulant are likelymost marked on the first few milliliters of blood collected.We hypothesize that these conditions result in Hb S po-lymerization in at least the first portion of blood col-lected, though it is not clear why this lesion does notrapidly reverse.

Further evidence for citrate solution-induced Hb Spolymerization in Hb AS blood was provided by the dif-ference in chemistries and RBC indices in CP2D compo-nents collected by phlebotomy between Hb As and Hb AAdonors. Sodium, potassium, chloride, glucose, osmolaritylevels, and pH differed in CP2D blood collected from HbAS and Hb AA donors. We speculate that these differ-

ences are due to the polymerization of Hb S. The fact thatno difference appeared in osmolarity or glucose levels inblood collected in heparin supports the conclusion thatCP2D collection lesion is responsible for Hb S polymer-ization.

We found that apheresis RBC components collectedfrom Hb AS donors filtered more effectively than RBCcomponents collected by phlebotomy into CP2D. Theimproved filterability of apheresis RBC components islikely due to avoidance of the citrate collection lesion. Incontrast to phlebotomy-collected CP2D components, nodifferences in the properties of apheresis componentscollected from Hb AS and Hb AA donors were noted,indicating that polymerization was less problematic inapheresis components. The reduction or elimination ofthe citrate collection lesion in apheresis RBC componentsis likely due to the method of addition of CP2D. Duringapheresis, CP2D is added to the blood at a carefully con-trolled rate proportional to the whole-blood collectionrate. In contrast, blood collected by phlebotomy flowsinto a bag containing enough CP2D to anticoagulate anentire unit.

Factors other than citrate collection lesions alsocontributed to Hb S polymerization. Although filtrationof apheresis RBC components was superior to phle-botomy components, some apheresis componentsdid not filter effectively. These components had lower Hboxygen saturations levels than those that filtered com-pletely. It is unclear why Hb oxygen saturation levelsvary among donors.

Blood from Hb AS donors collected in heparin fil-tered effectively, but heparin is not a suitable anticoagu-lant. An alternative to collecting blood in CP2D is to col-lect it in another citrate-based anticoagulant such as CPDor CPDA-1. The osmolarities of CPD and CPDA-1 are lessthan that of CP2D because they contain less dextrose.However, CPD and CPDA-1 are still hyperosmotic andacidic,19 and RBC filter performance problems have beenreported with RBC components from Hb AS donors col-lected in CPDA-1.5,20

These studies have several implications. Because fil-ter failures were due to changes in the rheologic proper-ties of the Hb AS RBCs, it would be expected that withoutsome intervention to prevent Hb S polymerization, RBCcomponents from Hb AS donors would occlude most, ifnot all, WBC-reduction filters. The incidence of failuremay vary among filters depending on the construction ofthe filter and filter material, the type of citrate anticoagu-lant, the temperature of the blood at the time of filtration,and the interval of storage before blood was filtered. It islikely that RBC components stored in gas-permeablebags before filtration would have increased oxygen ten-sions and would filter more effectively.

Our data show that when blood from donors with HbAS is collected under the appropriate conditions, filters

Fig. 1. Hb oxygen saturation is greater in Hb AS donor

apheresis RBC components that filter effectively. Hb oxygen

saturation was greater among Hb AS donors whose apher-

esis RBC components filtered completely in less than 15

minutes (black bars) than in Hb AS donors whose compo-

nents did not filter completely or required 100 or more min-

utes to filter (gray bars).

Hb S POLYMERIZATION CAUSES RBC FILTER FAILURES

Volume 42, November 2002 TRANSFUSION 1471

can be used to remove WBCs from RBC components. It isimportant to use existing collection systems, such asapheresis, or to develop new systems that permit WBCreduction of RBC components from Hb AS donors. Hb ASis most prevalent in African Americans, and because thispopulation is underrepresented among blood donors,high priority should be given to future efforts to adaptRBC collection techniques such that donors with Hb ASare not excluded from blood donation.

ACKNOWLEDGMENTS

The authors thank the staff of the Blood Services Section of

the Department of Transfusion for collecting blood from Hb

AS and Hb AA donors.

REFERENCES

1. Dzik S, Aubuchon J, Jeffries L, et al. Leukocyte reduction

of blood components: public policy and new technology.

Tranfus Med Rev 2000;14:34-52.

2. Kadar JG. Leukocyte-reduced blood products. Curr Opin

Hematol 1998;5:381-5.

3. Pietersz RN, van der Meer PF, Seghatchian MJ. Update on

leucocyte depletion of blood components by filtration.

Transfus Sci 1998;19:321-8.

4. Walker TW, Izaguirre C, Jenkins C, Sher G, Tittley P. Uni-

versal leukoreduction: Canadian experiences. Presented at

the Blood Products Advisory Committee meeting, June 15,

2001, Gaithersburg, MD.

5. Beard MJ, Seghatchian J, Cardigan R, Bennett J, Smith

KM, Williamson L. Leucofiltration of sickle cell trait blood

‘the blocker.’ NBS London S E Experience Transfus Sci

2000;22:71-3.

6. Ould Amar AK, Cesaire R, Kerob-Bauchet B, Robert P,

Maier H, Bucher B. Altered filterability of fresh sickle cell

trait donor blood (letter). Vox Sang 1997;73:55-6.

7. Gorlin JB, Adams CL, Stefan MM, Averbeck DA, Gonnella

KC. Variable leukocyte reduction on units from donors

with sickle cell trait: a time-temperature, donor reproduc-

ibility study (abstract). Transfusion 2000;40(Suppl):55.

8. Tinmouth AT, Naiman R, Garel C, Clark B, Asenbryl E,

Hannach B. Sickle trait associated with filter obstruction

during pre-storage leukodepletion of red blood cell prod-

ucts (abstract). Transfusion 2000;40(Suppl):59.

9. Beutler E. The sickle cell disease and related disorders. In:

Beutler E, Lichtman MA, Coller BS, Kipps TJ, eds.

Williams hematology. 5th ed. New York: McGraw-Hill,

1995:616-54.

10. Bunn HF. Pathogenesis and treatment of sickle cell dis-

ease. N Engl J Med 1997;337:762-9.

11. Steinberg MH. Management of sickle cell disease. N Engl

J Med 1999;340:1021-30.

12. Noguchi CT, Torchia DA, Schechter AN. Polymerization of

hemoglobin in sickle trait erythrocytes and lysates. J Biol

Chem 1981;256:4168-71.

13. Noguchi CT. Polymerization in erythrocytes containing S

and non-S hemogloblins. Biophys J 1984;45:1153-8.

14. Helzlsouer KJ, Hayden FG, Rogol AD. Severe metabolic

complications in a cross-country runner with sickle cell

trait. JAMA 1983;249:777-9.

15. Nuss R, Loehr JP, Daberkow E, Graham L, Lane PA. Car-

diopulmonary function in men with sickle cell trait at

moderately high altitude. J Lab Clin Med 1993;122:382-7.

16. Gupta AK, Kirchner KA, Nicholson R, Adams JG III,

Schechter AN, Noguchi CT, Steinberg MH. Effects of

alpha-thalassemia and sickle polymerization tendency

on the urine-concentrating defect of individuals with

sickle cell trait. J Clin Invest 1991;88:1963-8.

17. Gibson JG, Murphy WP, Scheitlin WA, Rees SB. The influ-

ence of extracellular factors involved with the collection

of blood in ACD on maintenance of red cell viability dur-

ing refrigerated storage. Am J Clin Pathol 1956;26:855-73.

18. Hiruma H, Noguchi CT, Uyesaka N, Schechter AN,

Rodgers GP. Contributions of sickle hemoglobin polymer

and sickle cell membranes to impaired filterability. Am J

Physiol 1995;268:H2003-8.

19. Vengelen-Tyler V, ed. Technical manual. 13th ed.

Bethesda, MD: AABB, 1999.

20. Kline L, George G, Luban NLC, Kuyer F. Leukoreduction

of red cells from donors with sickle cell trait. Transfusion

2001;41(Suppl):64.

STRONCEK ET AL.

1472 TRANSFUSION Volume 42, November 2002