Biol Blood Marrow Transplant 21 (2015) 688e695

Biology of Blood andMarrow Transplantationjournal homepage: www.bbmt.org

Effect of Cord Blood Processing on Transplantation Outcomesafter Single Myeloablative Umbilical Cord BloodTransplantation

Karen K. Ballen 1, Brent R. Logan 2,3, Mary J. Laughlin 4, Wensheng He 2, Daniel R. Ambruso 5,Susan E. Armitage 6, Rachel L. Beddard 7, Deepika Bhatla 8, William Y.K. Hwang 9, Joseph E. Kiss 10,Gesine Koegler 11, Joanne Kurtzberg 12,13, Arnon Nagler 14,15, David Oh 16, Lawrence D. Petz 17,Thomas H. Price 18, Ralph R. Quinones 19, Voravit Ratanatharathorn 20,21, J. Douglas Rizzo 2,Kathleen Sazama 22, Andromachi Scaradavou 23, Michael W. Schuster 24,25, Leonard S. Sender 26,Elizabeth J. Shpall 6, Stephen R. Spellman 27, Millicent Sutton 28, Lee Ann Weitekamp 29,John R. Wingard 22,30, Mary Eapen 2,*

1Department of Hematology/Oncology, Massachusetts General Hospital, Boston, Massachusetts2Department of Medicine, Center for International Blood and Marrow Transplant Research, Medical College of Wisconsin, Milwaukee, Wisconsin3Division of Biostatistics, Institute for Health and Society, Medical College of Wisconsin, Milwaukee, Wisconsin4Cleveland Cord Blood Center, Cleveland, Ohio5Center for Cancer and Blood Disorders, Children’s Hospital Colorado, Aurora, Colorado6MD Anderson Cord Blood Bank, Houston, Texas7 South Texas Blood and Tissue Center, San Antonio, Texas8 St.Louis Cord Blood Bank, St. Louis, Missouri9 Singapore Cord Blood Bank, Ltd., Singapore10 The Institute for Transfusion Medicine, Pittsburgh, Pennsylvania11Dusseldorf Cord Blood Bank, Dusseldorf, Germany12Division of Pediatric Blood and Marrow Transplantation, Department of Pediatrics, Duke University Medical Center, Durham, North Carolina13Carolinas Cord Blood Bank, Durham, North Carolina14Division of Internal Medicine, The Bone Marrow Transplantation Department, Sheba Medical Center, Tel-Hashomer, Israel15 Sheba Cord Blood Bank, Ramat Gan, Israel16 San Diego Blood Bank, San Diego, California17 StemCyte, Covina, California18 Puget Sound Blood Center, Seattle, Washington19University of Colorado Cord Blood Bank, Aurora, Colorado20Department of Clinical Arrangements, Karmanos Cancer Institute, Wayne State University, Detroit, Michigan21 J.P. McCarthy Cord Stem Cell Bank, Detroit, Michigan22 LifeSouth Community Blood Centers, Gainesville, Florida23New York Blood Center, National Cord Blood Program, Long Island City, New York24Department of Medicine, Stony Brook University Medical Center, Stony Brook, New York25Gift of Life Bone Marrow Foundation, Boca Raton, Florida26Children’s Hospital of Orange County Cord Blood Bank, Orange, California27Center for International Blood and Marrow Transplant Research, National Marrow Donor Program, Minneapolis, Minnesota28Community Blood Services, Camden, New Jersey29Michigan Blood Cord Blood Bank, Grand Rapids, Michigan30Divison of Hematology and Oncology, Department of Medicine, Shands HealthCare and University of Florida, Gainesville, Florida

Article history:

Received 7 August 2014Accepted 17 December 2014

Financial disclosure: See Acknowl* Correspondence and reprint re

for International Blood and Marrow

http://dx.doi.org/10.1016/j.bbmt.201083-8791/� 2015 American Socie

a b s t r a c tVariations in cord blood manufacturing and administration are common, and the optimal practice is notknown. We compared processing and banking practices at 16 public cord blood banks (CBB) in the UnitedStates and assessed transplantation outcomes on 530 single umbilical cord blood (UCB) myeloablativetransplantations for hematologic malignancies facilitated by these banks. UCB banking practices were

edgments on page 694.quests: Mary Eapen, MBBS, MS, CenterTransplant Research, Medical College of

Wisconsin, 9200 West Wisconsin Avenue, Suite C5500, Milwaukee,WI 53226.

E-mail address: meapen@mcw.edu (M. Eapen).

14.12.017ty for Blood and Marrow Transplantation.

K.K. Ballen et al. / Biol Blood Marrow Transplant 21 (2015) 688e695 689

Key Words:Cord blood bankProcessing methodsSurvival

separated into 3 mutually exclusive groups based on whether processing was automated or manual, unitswere plasma and red blood cell reduced, or buffy coat production method or plasma reduced. Compared withthe automated processing system for units, the day 28 neutrophil recovery was significantly lower aftertransplantation of units that were manually processed and plasma reduced (red cell replete) (odds ratio, .19;P ¼ .001) or plasma and red cell reduced (odds ratio, .54; P ¼ .05). Day 100 survival did not differ by CBB.However, day 100 survival was better with units that were thawed with the dextran-albumin wash methodcompared with the “no wash” or “dilution only” techniques (odds ratio, 1.82; P ¼ .04). In conclusion, CBBprocessing has no significant effect on early (day 100) survival despite differences in kinetics of neutrophilrecovery.

� 2015 American Society for Blood and Marrow Transplantation.

INTRODUCTIONUmbilical cord blood (UCB) transplantation has extended

access to hematopoietic stem cell transplantation (HCT) to adiverse racial and ethnic population [1]. Recent data havesuggested comparable results between UCB and grafts frommatched adult unrelated donors in both the myeloablativeand reduced-intensity transplantation setting [2-5]. How-ever, unlike bone marrow or peripheral blood, UCB units arecollected, cryopreserved, and, when needed, thawed andinfused.

Although the Food and Drug Administration has issuedguidance for manufacturing of cord blood for banking forunrelated transplantation, and the American Association ofBlood Banks and the Foundation for Accreditation ofCellular Therapy have established standards for productmanufacturing, practices at individual cord blood banks(CBB) vary tremendously. For example, UCB can either becollected in utero by trained obstetrical personnel and/or exutero by trained staff of the UCB bank. The American RedCross reported no difference in total nucleated cell (TNC)count, postprocessing CD34þ, or colony-forming units be-tween the 2 methods, but transplantation outcomes werenot assessed [6]. Similarly, processing of the UCB unit varieswidely among and within the CBBs. In the earliest years ofUCB banking, CBBs did not manipulate the product, otherthan diluting and adding dimethyl sulfoxide (DMSO), beforefreezing [7]. Today, most CBBs employ some form of volumereduction, which is generally achieved by depleting redblood cells, plasma, or both [8]. Each CBB has its own pro-cedures, some of which may have evolved over the history ofthe bank. Most CBBs have adopted the plasma and red bloodcell reduction method [9]. An alternative method is todeplete plasma but not red blood cells so that entrapment ofnucleated cells, and possibly progenitor and stem cells, isavoided, with some degree of volume reduction associatedwith the removal of plasma [10,11].

Appropriate handling and thawing of UCB units attransplantation centers are equally important to successfultransplantation outcomes. Pablo Rubinstein described athawing procedure using a dextran and albumin solution toremove DMSO. The majority of transplantation centersadopted this approach, and nucleated cell count recoveries of75% to 90% have been reported [12,13]. More recently, Barkeret al. described a dilution-only “no wash” method withreconstitution in dextran-albumin for a final 5% DMSO con-centration [14]. Nucleated cell count recovery was 86%, andthere were no serious adverse infusion events. Finally, somecenters have used a nonvolume-reduced (unmanipulated)thawing strategy and have demonstrated adequate engraft-ment [15]. The report of several life-threatening infusionreactions with UCB infusion have intensified the need todetermine the optimal thawing practice [16].

The optimal processing techniques for UCB units are notestablished and whether transplantation outcomes differ bytechniques is not clear. Therefore, we collected informationon UCB processing at the CBBs and examined the data for aneffect of processing methods at CBBs in patients who hadundergone a single UCB transplantation for acute leukemiaor myelodysplastic syndrome, the most common indicationsfor allogeneic HCT. This report, the first of its kind, providesadditional knowledge on whether practices at CBBs tech-niques influence hematopoietic recovery and early survivalafter UCB transplantation.

MATERIALS AND METHODSData Source

The Center for International Blood and Marrow Transplant Research(CIBMTR) is a working group of over 450 transplantation centers worldwidethat contribute detailed data on consecutive allogeneic and autologous HCTto a statistical center at the Medical College of Wisconsin or the data coor-dinating center at the National MarrowDonor Program. Banking practices atCBBs were obtained using a short survey, which addressed UCB unit pro-cessing at the banks. Data on UCB unit thawing at transplantation centerswere obtained through standardized data collection forms developed by theCIBMTR. Patients provide written informed consent for participation inaccordance with the Declaration of Helsinki. The institutional review boardsof the Medical College of Wisconsin and the National Marrow Donor Pro-gram approved the study.

PatientsIncluded are 530 patients with acute myeloid leukemia (AML), acute

lymphoid leukemia or myelodysplasia who received single unit unrelatedUCB transplantation in the United States with a UCB unit from 1 of the 16participating CBBs. All transplantations occurred in the United States be-tween the years 2000 and 2011. Only recipients of myeloablative regimens,defined as having received total body irradiation dose of 1000 cGy or higheror busulfan dose greater than 9 mg/kg or melphalan dose greater than150 mg/m2, are included [17]. Recipients of multiple or expanded UCB units,reduced-intensity conditioning regimens, and transplantations for nonma-lignant diseases were excluded.

CBB PracticesSixteen publically funded CBBs in the United States participated in the

survey. Using banking practices reported, 3 mutually exclusive groups werecreated (Table 1) based on the following: automated ormanual processing atthe CBB andwhether units were plasma and red blood cell reduced, used thebuffy coat production method, or were plasma reduced. All units containedDMSO and an hyperosmolar agent. Group 1 included units that were pro-cessed using an automated method that were plasma and red blood cellreduced (n¼ 84) or used the buffy coat productionmethod (n¼ 34). Group 2included manually processed units that were plasma and red blood cellreduced (n¼ 274) or used the buffy coat productionmethod (n¼ 5). Group 3includedmanually processed units that were plasma reduced. Of note, as thegroups were created based on self-reported practices, some CBBs are rep-resented in more than 1 group, as banking practices evolved over the studyperiod. Further, group 3 represents a single CBB and the buffy coat pro-duction method is implemented at 3 CBBs and represented in group 1(n ¼ 34 from a single CBB) and group 2 (n ¼ 5 from 2 CBBs).

OutcomesThe primary endpoint was hematopoietic recovery; neutrophil recovery

was defined as achieving an absolute neutrophil count � .5 � 109/L for 3

Table 1Cord Bank Practices

Characteristic Group 1 Group 2 Group 3

Practices Automated processing, plasmaand red cell reduced or buffycoat production method

Manual processing, plasma andred cell reduced or buffy coatproduction method

Manual processing,plasma reduction

No. of banks* 6 13 1Type of processing systemAutomated 6Manual 13 1

Product processing method at banksPlasma and red cell reduction 5 11Buffy coat production 1 2Plasma reduction 1

AnticoagulantACD 1CPD 4 12 1CPDA 1 1

Storage method at bankVapor phase 1 6 1Liquid nitrogen 5 8

Years of existence of bank�10 yr 2 211-15 yr 2 7 115-20 yr 2 4

ACD indicates acid citrate dextrose; CPD, citrate phosphate dextrose; CPDA, citrate phosphate adenine.* Number of banks exceeds N ¼ 16 because practice changes occurred over time and those banks are represented in the Table more than once.

K.K. Ballen et al. / Biol Blood Marrow Transplant 21 (2015) 688e695690

consecutive days and platelet recovery as achieving platelets � 20 � 109/L,unsupported for 7 days. Death from any cause was considered an event.

Statistical MethodsThe characteristics of patients, their disease, and transplantation

grouped according to CBB practice were compared using the chi-square testfor categorical variables [18]. The incidences of neutrophil and platelet re-covery were calculated using the cumulative incidence estimator; deathwithout an event was considered a competing risk [19]. The day 100probability of overall survival was calculated using the Kaplan-Meier esti-mator [18]. Generalized linear mixed models were used to fit random effectlogistic regression models for day 28 and day 42 neutrophil, day 90 plateletrecovery, and day 100 overall survival. These models included random ef-fects for CBB and for transplantation center to account for within bank orwithin center correlation when examining CBB-specific variables or trans-plantation centerespecific variables [20]. The following variables weretested in a multivariate analysis: CBB practice groups (1, 2, and 3) and itseffect on outcomes was the primary interest. Consequently, the variable forCBB practice group was held in all steps of model building, regardless of thelevel of significance attained. Other CBB variables tested included vaporversus liquid storage, year of collection (1996 through 1999 versus 2000through 2005 versus 2006 through 2011), and length of existence of the CBB(�10 versus 11 through 15 versus > 15 years). Transplantation centerpractice was tested as follows: no DMSO dilution/no wash versus reconsti-tution in dextran/plasmalyte A and interval between thaw and completionof infusion (<2 versus 2 through 4 versus > 4 hours). In addition to CBBpractice and transplantation center practices, patient (age, gender, perfor-mance score, cytomegalovirus serostatus, race), disease, disease status attransplantation, and transplantation characteristics (conditioning regimen,in vivo Tcell depletion, graft-versus-host disease prophylaxis, HLAmatch [6/6 versus 5/6 versus 4/6], TNC dose [< 3 versus 3 to 5 versus > 5 � 107/kg],donor-recipient race match [donor/recipient same race versus other], in-terval between unit collection and transplantation [< 1 versus 1 through 3versus > 3 through 5 versus > 5 years], and transplantation period [2000through 2004 versus 2005 through 2011]) were tested in all multivariatemodels such that the effects of CBB and transplantation center practice wereadjusted for known clinical characteristics associated with outcome. A Pvalue of .05 or less was considered significant; all P values were 2-sided.Analyses were performed with SAS 9.3 (Cary, NC).

RESULTSPatient, Disease, and Transplantation Characteristics

Table 2 shows the characteristics of the 530 patientsgrouped by processing methods at the CBBs. Overall, thecharacteristics of patients, their diseases, and trans-plantations were similar across the 3 groups except for

patient race, TNC, thaw procedures at transplantation cen-ters, transplantation conditioning regimen, in vivo T celldepletion, planned growth factor treatment, and trans-plantation period. Caucasians were less likely to havereceived units that were manually processed and plasmareduced. The manually processed units were more likely tocontain prefreeze TNC greater than 5�107/kg but there wereno differences in post-thaw TNC recovery between the 3groups. During the thawing process, the automated pro-cessed units, which account for the more recent trans-plantations, were less likely to have been washed andreconstituted in dextran. Total body irradiation trans-plantation conditioning regimens were more commonlyused in patients who received units that were processedusing an automated technique. On the other hand, in vivo Tcell depletion was common with manually processed unitsand reflect clinical practice in an earlier period. Plannedgrowth factor therapy was less common for units that weremanually processed and plasma reduced.

Neutrophil and Platelet RecoveryThe primary endpoint of the study was hematopoietic

recovery: day 28 neutrophil recovery and day 90 plateletrecovery (Table 3). The likelihood of neutrophil recovery waslower after transplantation of UCB units that were processedmanually and plasma reduced (red blood cell replete) orplasma and red cell depleted (Table 3, Figure 1). Further,neutrophil recovery was lower after transplantation ofmanually processed units with plasma reduction comparedwith plasma and red cell reduction (odds ratio, .34; 95%confidence interval [CI], .16 to .74; P ¼ .04). The day 28 cu-mulative incidence of neutrophil recovery was 40% (95% CI,32% to 48%) after transplantation of UCB units that wereprocessed manually and plasma reduced; median time torecovery was 29 days. Corresponding probabilities for unitsthat were processed using automated and manual methodswith plasma and red blood cell reduction or buffy coat pro-duction method were 77% (95% CI, 69% to 84%) and 68% (95%CI, 63% to 73%), respectively; median time to neutrophil

Table 2Patients, Disease, Transplantation, and Transplantation Center Practices

Variable Group 1 Group 2 Group 3 P Value

Description Automated processing, plasmaand red cell reduced or buffycoat production method

Manual processing, plasma andred cell reduced or buffy coatproduction method

Manual processing,plasma reduced

No. of patients 118 279 133Age .47�16 yr 89 (75) 225 (81) 107 (80)>16 yr 29 (25) 54 (19) 26 (20)

Sex .76Male 55 (47) 141 (51) 67 (50)Female 63 (53) 138 (46) 66 (50)

Performance score .21<90 20 (17) 51 (18) 33 (25)90-100 97 (82) 218 (78) 97 (73)Not reported 1 (<1) 10 (4) 3 (2)

Recipient CMV serostatus .06Negative 46 (39) 137 (49) 53 (40)Positive 71 (60) 137 (49) 79 (59)Not reported 1 (<1) 5 (1) 1 (1)

Recipient race <.001Caucasian 61 (52) 181 (65) 55 (41)Non-Caucasian 57 (48) 92 (33) 77 (58)Not reported 0 6 (2) 1 (<1)

Disease status at transplantation .02CR1/RA 52 (44) 92 (33) 53 (40)CR2 53 (45) 127 (46) 48 (36)Relapse/RAEB 13 (11) 60 (21) 32 (24)

Conditioning regimen .04TBI-containing 96 (81) 193 (69) 100 (75)Non-TBI containing 22 (19) 86 (31) 33 (25)

GVHD prophylaxis .26Tacrolimus-containing 33 (28) 77 (28) 47 (35)Cyclosporine-containing 83 (70) 199 (71) 85 (64)Other 2 (2) 3 (1) 1 (1)

HLA match low resolution at A, B, andallele-level DRB1

.42

4/6 42 (36) 88 (32) 55 (41)5/6 57 (48) 141 (51) 58 (44)6/6 19 (16) 50 (17) 20 (15)

Donor-recipient race match .02Donor recipient same race 71 (60) 150 (54) 70 (53)Donor recipient different race 47 (40) 72 (26) 62 (47)Unknown 0 57 (20) 1 (1)

In vivo T depletion <.001No 76 (64) 116 (42) 58 (44)Yes 42 (36) 163 (58) 75 (56)

TNC dose (107/kg)�3.0 11 (9) 32 (11) 17 (13) .013.0-5.0 44 (37) 74 (27) 24 (18)>5.0 61 (52) 172 (62) 92 (69)Not reported 2 (2) 1 (<1) 0

Post-thaw TNC recovery .77<70% 21 (18) 56 (20) 27 (20)70%-90% 53 (45) 120 (43) 62 (47)>90% 23 (19) 67 (24) 28 (21)Not reported 21 (18) 36 (13) 16 (12)

Thawing method at transplantation center <.001No wash/DMSO dilution 44 (37) 42 (15) 24 (18)Washed, reconstituted in dextran/hespan/

plasmalyte71 (60) 219 (78) 95 (71)

Not reported 3 (3) 17 (7) 14 (11)Interval between thaw to completing of

infusion.62

<2 hr 38 (32) 91 (33) 45 (34)2-4 hr 56 (47) 127 (46) 56 (42)>4 hr 11 (9) 32 (11) 22 (17)Not reported 13 (11) 29 (10) 10 (8)

Interval from unit collection totransplantation

<.001

�1 yr 27 (23) 22 (8) 21 (16)1-3 yr 64 (54) 97 (35) 48 (36)3-5 yr 19 (16) 76 (27) 33 (25)>5 yr 8 (7) 84 (30) 31 (23)

(Continued on next page)

K.K. Ballen et al. / Biol Blood Marrow Transplant 21 (2015) 688e695 691

Table 2(continued)

Variable Group 1 Group 2 Group 3 P Value

Planned growth factor therapy .02None 34 (29) 87 (31) 60 (45)Yes 68 (58) 152 (54) 52 (39)Not reported 16 (14) 40 (15) 21 (16)

Year of transplantation <.0012000-2004 6 (5) 57 (20) 21 (16)2005-2011 112 (95) 222 (80) 112 (84)

CMV indicates cytomegalovirus; CR, complete remission; RA, refractory anemia; RAEB, refractory anemia with excess blasts; TBI, total body irradiation; GVHD,graft-versus-host disease.

K.K. Ballen et al. / Biol Blood Marrow Transplant 21 (2015) 688e695692

recovery was 20 and 21 days, respectively. Independent ofprocessing methods, the likelihood of neutrophil recoverywas higher after transplantation of units with TNC greaterthan 5 � 107/kg compared with units with TNC < 3 � 107/kgand older patients (Table 3). Despite early differences in re-covery rates, by 6 weeks after transplantation, there were noappreciable differences between the 3 groups (Table 3).

There was no difference in platelet recovery at day 90among the 3 groups (Table 3). The day 90 probabilities of

Table 3Risk Factors Associated with Neutrophil Recovery at Day 28 and Day 42 Platelet R

Risk Factor N1/N2

Neutrophil recovery at day 28Banking practice*

Group 1 92/116Group 2 188/270Group 3 60/132

Age�16 yr 269/412>16 yr 71/106

TNC dose, � 107/kg�3 31/58>3-5 89/140>5 220/320

Neutrophil recovery at day 42Group 1 104/116Group 2 225/270Group 3 108/132

TNC dose, � 107/kg�3 40/58>3-5 117/140>5 280/320

Conditioning regimenTBI þ cyclophosphamide 327/379Busulfan þ cyclophosphamide 110/139

Platelet recovery at day 90Group 1 76/116Group 2 171/270Group 3 69/132

GenderFemale 145/260Male 171/258

TNC dose, � 107/kg�3 23/58>3-5 88/140>5 205/320

Disease statusy

Early 128/193Intermediate 141/222Advanced 47/103

N1 indicates number of events; N2, number evaluable.Odds ratios>1 indicate better outcome. Other comparisons include day 28 neutropday 42 neutrophil recovery (group 3 versus group 2; OR, .81 [95% CI, .47 to 1.42]

* Group 1: automated processing, units were plasma and red blood cell reducedplasma and red blood cell reduced or buffy coat production method. Group 3: m

y Early risk includes patients in first complete remission/refractory anemia. Intincludes patients in relapse/refractory anemia with excess blasts.

platelet recovery were 65% (95% CI, 57% to 74%), 64% (95% CI,58% to 69%), and 53% (95% CI, 44% to 61%) for groups 1, 2, and3 respectively. The median time to platelet recovery was 46,48, and 61 days, respectively. Additionally, platelet recoverywas better after transplantation of units with TNC greaterthan 5 � 107/kg compared with units with TNC 3 � 107/kg orlower and in males compared with females. Platelet recoverywas lower in patients who underwent transplantation withactive disease compared with those who underwent

ecovery at Day 90

Odds Ratio (95% CI) P Value

1.00.54 (.29-1.00) .05.19 (.08-.44) <.001

1.002.01 (1.11-3.67) .02

1.001.65 (.82-3.31) .163.26 (1.57-6.77) .002

1.00.61 (.31-1.21) .16.50 (.23-1.06) .07

1.002.11 (1.03-4.33) .043.74 (1.90-7.35) <.001

1.00.52 (.30-.88) .02

1.00.98 (.61-1.58) .93.61 (.36-1.05) .08

1.001.58 (1.09-2.28) .02

1.002.43 (1.27-4.66) .0082.64 (1.46-4.78) .002

1.00.82 (.54-1.25) .36.45 (.27-.75) .002

hil recovery (group 3 versus group 2; OR, .34 [95% CI, .16 to .74], P¼ .04) and, P ¼ .46).or buffy coat production method. Group 2: manual processing, units were

anual processing, units were plasma reduced.ermediate risk includes patients in second complete remission. Advanced

100

0

20

40

60

80

Pro

babi

lity,

%

Months0 2 431

Group 1

Group 2

Group 3

Figure 2. Overall survival. Group 1: automated processing, units were plasmaand red blood cell reduced or buffy coat production method. Group 2: manualprocessing, units were plasma and red blood cell reduced or buffy coat pro-duction method. Group 3: manual processing, units were plasma reduced.

100

0

20

40

60

80

Cum

ulat

ive

Inci

denc

e, %

Days

Group 1

Group 2

Group 3

0 14 4228 35217

Figure 1. Neutrophil recovery. Group 1: automated processing, units wereplasma and red blood cell reduced or buffy coat production method. Group 2:manual processing, units were plasma and red blood cell reduced or buffy coatproduction method. Group 3: manual processing, units were plasma reduced.

K.K. Ballen et al. / Biol Blood Marrow Transplant 21 (2015) 688e695 693

transplantation in first remission. The effect of TNC anddisease status on platelet recovery was independent of CBBprocessing methods.

Of note, we carefully examined for an effect of duration ofstorage UCB units and its effect on hematopoietic recovery.Neutrophil recovery at day 28 and day 42 was not associatedwith duration of storage (P ¼ .87 and P ¼ .51, respectively).Similarly, platelet recovery at day 90 was also not associatedwith duration of storage of UCB units (P ¼ .89).

Overall SurvivalDespite a lower threshold of early neutrophil recovery

after transplantation of manually processed units, therewereno differences among the groups with respect to overallsurvival at day 100 (Table 4, Figure 2). The day 100 survivalprobabilities for groups 1, 2, and 3 were 83% (95% CI, 76% to

Table 4Risk factors Associated with Overall Survival at Day 100

Risk Factor N1/N2 Odds Ratio(95% CI)

P Value

Banking practice*

Group 1 96/116 1.00Group 2 218/270 .94 (.51-1.73) .85Group 3 106/132 .93 (.47-1.84) .84

GenderFemale 199/260 1.00Male 221/258 1.84 (1.16-2.91) .01

Disease statusy

Early 165/193 1.00Intermediate 183/222 .82 (.48-1.40) .47Advanced 72/103 .44 (.24-.80) .007

Unit processing attransplantation center

No wash/DMSO dilution 84/108 1.00Thaw/wash þ dextran

reconstitution314/376 1.82 (1.02-3.23) .04

Transplantation period2000-2004 61/832005-2011 359/435 2.06 (1.14-3.72) .02

N1 indicates number alive; N2, number evaluable.Odds ratios >1 indicate better outcome.

* Group 1: automated processing, units were plasma and red blood cellreduced or buffy coat production method. Group 2: manual processing,units were plasma and red blood cell reduced or buffy coat productionmethod. Group 3: manual processing, units were plasma reduced.

y Early risk includes patients in first complete remission/refractory ane-mia. Intermediate risk includes patients in second complete remission.Advanced risk includes patients in relapse/refractory anemia with excessblasts.

89%), 82% (95% CI, 74% to 88%), and 80% (95% CI, 73% to 87%),respectively. However, there were other factors associatedwith early survival that were independent of CBB processingmethods. Survival was higher for males, for transplantationsin patients in first complete remission/refractory anemia,and transplantations after 2004. Independent of CBB prac-tice, thawing practice at transplantation centers alsoinfluenced early survival; survival was higher after trans-plantation of units that were thawed, washed, and recon-stituted in dextran compared with units that were notwashed or underwent DMSO dilution only (Table 4). We didnot observe differences in early survival after transplantationof units that were not washed (n ¼ 40; 78% [95% CI, 63% to89%]) and DMSO dilution only (n ¼ 70; 77% [95% CI, 67%to 86%]).

Toxicities Associated with InfusionThere were no life-threatening or serious or suspected

serious adverse events as defined by the Food and DrugAdministration’s Safety Reporting Requirements (21 CFR312.32), reported for any of the transplantations in the cur-rent analysis. One of the infused units was infected, asdefined by a positive product culture (organism not re-ported). This patient achieved neutrophil and platelet re-covery and is alive, 2 years after transplantation.

DISCUSSIONIn this study, the first of its kind, we explored whether

differences exist in hematopoietic recovery and early sur-vival by processingmethods at the CBBs. The current analysiswas prompted by the concern that variations in processingand banking have never been fully studied and that thesepractice variations might affect the early outcomes of UCBtransplantation. In a carefully controlled analysis andadjusting for known patient, disease, and transplantationcharacteristics, we identified 2 banking practices withadverse effects on early neutrophil recovery (ie, at day 28).Neutrophil recovery was lower after transplantation ofmanually processed UCB units that were plasma reduced(red blood cell replete) or plasma and red cell depletedcompared with automated methods of processing withplasma and red cell depletion. Automated processing isrelatively recent, and only 118 UCB units were processedusing automated techniques in this study [21]. Further,among the manually processed units, neutrophil recoverywas less likely with plasma reduction. Lower rates of

K.K. Ballen et al. / Biol Blood Marrow Transplant 21 (2015) 688e695694

neutrophil recovery within 4 weeks of transplantation didnot translate into higher early overall mortality, implyingadvances in supportive care during the early neutropenicperiod in part negated its adverse effects on early survival.Additionally, thawing methods were associated with earlysurvival. Survival was better when units were washed andreconstituted in dextran, hespan, or plasmalyte, implyingthawing techniques at transplantation centers may influenceearly survival. The observed adverse effect of thawing tech-niques at transplantation centers was independent of unitprocessing and banking at CBBs; nor was it associated withhematopoietic recovery. The results of the bedside thawapproach in 26 transplant recipients and that from the thawand dilute approach in 54 patients from single institutionsdemonstrated these approaches were safe and with sus-tained hematopoietic recovery [14,15].

To our knowledge, the current analysis is the first to havecompared the 3 approaches to UCB thaw. It is worth notingthat we do not have detailed information on handling andprocessing of UCB units at the transplantation centers orinformation as to whether the transplantation centers fol-lowed the recommendations for thawing by the CBB, all ofwhich may have an effect on survival. With the data avail-able, it is not possible to recommend 1 approach overanother other than to recommend UCB unit thaw that eitheradheres to the recommendations from the CBB or adopttechniques that have been validated by the institution’s cellprocessing laboratory. We did not identifiy differences inhematopoietic recovery or overall survival based on the yearsof banking experience, year of collection of UCB unit, storagecondition (vapor versus liquid phase), and the length of timethe UCB unit was stored at the CBB.

Our findings differ from that reported by others on theeffects of transplantation of UCB units that were plasmareduced [11]. The report by Chow et al. [11] demonstratedneutrophil recovery rates higher than reported in the currentanalysis. An important difference between that analysis andthe current analysis is that in the former, neutrophil recoveryrates were calculated with the Kaplan-Meier method,whereas the current analysis used the cumulative incidenceestimator, which considers the competing risk for the eventof interest, ie, death without recovery.

As expected from other studies in the literature, day 100survival was improved for patients who underwent trans-plantation in first complete remision and for those patientswho underwent transplantation after 2004 [22,23].Although higher performance status, younger age, andhigher cell dose were associated with improved survival inother larger studies on single UCB transplantation, thesefactors did not affect survival in this current study and werelikely attributed to that fact that we explored factors asso-ciated with survival within 3 months after transplantation[24]. The only plausible explanations include the fact that thecurrent analysis is limited to primarily a pediatric cohort,with about 80% of patients ages � 16 years and that theirperformance scores were 90 or 100. Similarly, almost 90% ofpatients received UCB units with a TNC dose in excess of3 � 107/kg.

The current analysis has several limitations. First, ouranalysis is limited to approximately 500 and largely, pedi-atric, recipients of single unit UCB transplantation andcomprises approximately 10% of UCB units distributed by theparticipating CBBs. To allow us to carefully examine for CBBprocessing and banking, we limited the cohort to a relativelyhomogenous group of patients who received a single UCB

unit, which explains the relatively small numbers of trans-plant recipients in the current analysis; we arbitrarilyassigned CBB practices into 3 broad groups. Second, althoughwe adjusted for known prognostic factors, there may beseveral unknown and unmeasured practices both at CBBsand transplantation centers that influenced the outcome.Third, wewere not able to study storage temperature at CBBs[25], cell viability, or the cord blood Apgar score [26]. Fourth,CBBs were surveyed for their banking practice with relianceon self reporting. All other data were collected on stan-dardized CIBMTR reporting forms with appropriate datamanagement manuals and subject to audit.

Despite the limitations, our observations have clinicalrelevance. Banking practices are not associated with earlysurvival, but manual methods of UCB unit processing lowerthe odds of neutrophil recovery during the very early post-transplantation period. Although only 6 CBBs were usingautomated processing methods at time of survey about ayear ago, it is possible several CBBs have switched to theautomated methods. These data support selecting anadequately dosed and HLA-matched or HLA-mismatchedUCB unit that was processed using automated methodswhen such a unit is available.

ACKNOWLEDGMENTThe CIBMTR is supported by Public Health Service Grant/

Cooperative Agreement U24-CA076518 from the NationalCancer Institute, the National Heart, Lung, and Blood Insti-tute and the National Institute of Allergy and InfectiousDiseases; a grant/cooperative agreement 5U10HL069294from NHLBI and NCI; a contract HHSH250201200016C withHealth Resources and Services Administration (HRSA/DHHS); 2 grants N00014-12-1-0142 and N00014-13-1-0039from the Office of Naval Research; and grants from *ActiniumPharmaceuticals; Allos Therapeutics, Inc.; *Amgen; anony-mous donation to the Medical College of Wisconsin; Ariad;Be The Match Foundation; *Blue Cross and Blue Shield As-sociation; *Celgene Corporation; Chimerix, Inc.; FredHutchinson Cancer Research Center; Fresenius-BiotechNorth America, Inc.; *Gamida Cell Teva Joint Venture Ltd.;Genentech, Inc.; *Gentium SpA; Genzyme Corporation;GlaxoSmithKline; Health Research, Inc.; Roswell Park CancerInstitute; HistoGenetics; Incyte Corporation; Jeff GordonChildren’s Foundation; Kiadis Pharma; Medac GmbH; TheMedical College of Wisconsin; Merck & Co., Inc.; Millennium:The Takeda Oncology Co.; *Milliman USA, Inc.; *MiltenyiBiotec; National Marrow Donor Program; Onyx Pharmaceu-ticals; Optum Healthcare Solutions, Inc.; Osiris Therapeutics;Otsuka America Pharmaceutical, Inc.; Perkin Elmer, Inc.;*Remedy Informatics; *Sanofi US; Seattle Genetics; Sigma-Tau Pharmaceuticals; Soligenix, Inc.; St. Baldrick’s Founda-tion; StemCyte, A Global Cord Blood Therapeutics Co.;Stemsoft Software, Inc.; Swedish Orphan Biovitrum; *TarixPharmaceuticals; *Terumo BCT; *Teva Neuroscience, Inc.;*Therakos; University of Minnesota; University of Utah; and*WellPoint The views expressed in this article do not reflectthe official policy or position of the National Institute ofHealth, the Department of the Navy, the Department of De-fense, Health Resources and Services Administration or anyother agency of the US Government.

*Corporate Members.Conflict of interest statement: There are no conflicts of

interest to report.Financial disclosure: The conduct of the study detailed

in the manuscript utilized data and resources from the

K.K. Ballen et al. / Biol Blood Marrow Transplant 21 (2015) 688e695 695

Center for International Blood and Marrow TransplantResearch.

REFERENCES1. Barker JN, Byam CE, Kernan NA, et al. Availability of cord blood extends

access to allogeneic hematopoietic stem cell transplant to racial andethnic minorities. Biol Blood Marrow Transplant. 2010;16:1541-1548.

2. Chen YB, Aldridge J, Kim HT, et al. Reduced-intensity conditioning stemcell transplantation: comparison of double umbilical cord blood andunrelated donor grafts. Biol Blood Marrow Transplant. 2012;18:805-812.

3. Brunstein CG, Barker JN, Wesidorf DJ, et al. Umbilical cord bloodtransplantation after nonmyeloablative conditioning: impact ontransplantation outcomes in 110 adults with hematologic disease.Blood. 2007;110:3064-3070.

4. Eapen M, Rocha V, Sanz G, et al. Effect of graft source on unrelateddonor hematopoietic stem-cell transplantation in adults with acuteleukemia: a retrospective analysis. Lancet Oncol. 2010;11:653-660.

5. Ballen KK, Spitzer TR, Yeap BY, et al. Double unrelated reduced-intensity umbilical cord blood transplantation in adults. Biol BloodMarrow Transplant. 2007;13:82-89.

6. Lasky LC, Lane T, Miller J, et al. In utero or ex utero cord bloodcollection: which is better? Transfusion. 2002;42:1261-1267.

7. Gluckman E, Broxmeyer HE, Auerbach AD, et al. Hematopoietic recon-stitution in a patient with Fanconi anemia by means of umbilical cordblood from a HLA-identical sibling. N Engl J Med. 1989;321:1174-1178.

8. Broxmeyer HE, Gordon GW, Hangoc G, et al. Human umbilical cordblood as a potential source of transplantable hematopoietic stem/progenitor cells. Proc Natl Acad Sci U S A. 1989;86:3828-3832.

9. Rubinstein P, Dobrila L, Rosenfield RE, et al. Processing and cryopres-ervation of placental/umbilical cord blood for unrelated bone marrowreconstiution. Proc Natl Acad Sci U S A. 1995;92:10119-10122.

10. Chow R, Nademanee M, Rosenthal J, et al. Analysis of hematopoieticcell transplants using plasma-depleted cord blood products that arenot red blood cell reduced. Biol Blood Marrow Transplant. 2007;13:1346-1357.

11. Chow R, Lin A, Tonai R, et al. Cell recovery comparison between plasmadepletion/reduction- and red cell reduction-processing of umbilicalcord blood. Cytotherapy. 2011;13:1105-1119.

12. Laroche V, McKenna DH, Moroff G, et al. Cell loss and recovery inumbilical cord blood processing: a comparison of postthaw and post-wash samples. Transfusion. 2005;45:1909-1916.

13. Zinno F, Landi F, Aureli V, et al. Pre-transplant manipulation process-ing of umbilical cord blood units: Efficacy of Rubinstein’s thawing

technique used in 40 transplantation procedures. Transfus Apher Sci.2010;43:173-178.

14. Barker JN, Abboud M, Rice RD, et al. A “No-wash” albumin-dextrandilution strategy for cord blood unit thaw: high rate of engraftmentand a low incidence of serious infusion reactions. Biol Blood MarrowTransplant. 2009;15:1596-1602.

15. Hahn T, Bunworasate U, George MC, et al. Use of no volume-reduced(unmanipulated after thawing) umbilical cord blood stem cells forallogeneic transplantation results in safe engraftment. Bone MarrowTransplant. 2003;32:145-150.

16. Choi S, Hoffmann S, Cooley L. Another case of acute cardiopulmonarytoxicity with cord blood infusion: is dextran the culprit? Transfusion.2012;52:207-208.

17. Bacigalupo A, Ballen KK, Rizzo D, et al. Defining the intensity of theconditioning regimen: working definitions. Biol Blood Marrow Trans-plant. 2009;15:1628-1633.

18. Klein JP, Moeschberger ML. Survival analysis: statistical methods forcensored and truncated data, 2nd ed. New York, NY: Springer-Verlag;2003.

19. Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failureprobabilities in the presence of competing risk: new representations ofold estimators. Stat Med. 1999;18:695-706.

20. Wolfinger R, O’Connell M. Generalized linear mixed models: a pseudo-likelihood approach. J Stat Comp Sim. 1993;48:233-243.

21. Lapierre V, Pellegrini N, Bradley I, et al. Cord blood volume reductionusing an automated system (Sepax) versus a semi-automated system(Optipress II) and a manual method (hydroxyethyl starch sedimenta-tion) for routine cord blood banking: a comparative study. Cytotherapy.2007;9:165-169.

22. Ballen KK, Gluckman E, Broxmeyer H. Umbilical cord blood trans-plantation: the first 25 years and beyond. Blood. 2013;122:491-498.

23. Peffault de Latour R, Brunstein CG, Porcher R, et al. Similar overallsurvival using sibling, unrelated donor and cord blood grafts afterreduced-intensity conditioning for older patients with acute myeloidleukemia. Biol Blood Marrow Transplant. 2013;19:1355-1360.

24. Ballen KK, Koreth J, Chen YB, et al. Selection of optimal alternative graftsource: mismatched unrelated donor, umbilical cord blood, or hap-loidentical transplant. Blood. 2012;119:1972-1980.

25. Louis I, Wagner E, Dieng MM, et al. Impact of storage temper-ature and processing delays on cord blood quality: discrepancybetween functional in vitro and in vivo assays. Transfusion. 2012;52:2401-2405.

26. Page KM, Zhang L, Mendizabal A, et al. The cord blood apgar: a novelscoring system to optimize selection of banked cord blood grafts fortransplantation. Transfusion. 2012;52:272-283.