peripheral blood stem cell yield calculated using preapheresis absolute cd34+ cell count, peripheral...

T R A N S P L A N T A T I O N A N D C E L L U L A R E N G I N E E R I N G

Peripheral blood stem cell yield calculated using preapheresisabsolute CD34+ cell count, peripheral blood volume processed,

and donor body weight accurately predicts actual yield atmultiple centers

Chitra Hosing,1 Rima M. Saliba,1 Nelson Hamerschlak,2 Jose Mauro Kutner,2

Araci Massami Sakashita,2 Andrea Tiemi Kondo,2 Morgani Rodrigues,2 Juliana Folloni Fernande,2

Alexandre Chiattone,1 Viviane C. Chiattone,1 Jose C. Barros,3 Carlos S. Chiattone,3

Ricardo Chiattone,3 Uday Popat,1 Muzaffar Qazilbash,1 Xiao Wen Tang,4 Depei Wu,4

Alejandro Majilis,5 Marcos de Lima,6 and Timoleon Anguita5

BACKGROUND: Accurate prediction of stem cell yieldis important for planning leukapheresis procedures. Aformula has been published (Pierelli et al., Vox Sang2006;91:126-34) to estimate the CD34+ dose collectedon the first day of leukapheresis that was based on thepreapheresis peripheral blood (PB) CD34+ counts,the blood volume processed, and the donor’s weight.The aim of this study was to assess the predictivevalue of this formula.STUDY DESIGN AND METHODS: Data were retro-spectively collected on 1126 consecutive PB stem cellharvests conducted at five institutions. Information onage, sex, diagnosis, weight, preapheresis absoluteperipheral CD34+ count, total blood volume processed,and CD34+ cells harvested per kilogram of body weighton the first day of apheresis was collected.RESULTS: Among donors at least 18 years old, Pear-son’s correlation coefficient (r) between actual yield(AY) and predicted yield (PY) was 0.76. To characterizethis correlation, AY and PY were classified as beingwithin the conventionally acceptable CD34+ doses(>2 × 106-5 × 106 cells/kg), below this range (≤2 × 106

cells/kg), or above it (>5 × 106 cells/kg). The positivepredictive value (PPV) of PY was estimated consideringthe distribution of AY as the “gold standard.” PPV wasrelatively high for PY of more than 5 × 106 cells/kg(85%), moderate for PY of not more than 2 × 106

cells/kg (72%), and low for PY more than 2 × 106 to5 × 106 cells/kg (56%). A consistent pattern wasobserved within institutions.CONCLUSION: The formula of Pierelli et al. is associ-ated with a PPV that is high, moderate, and relativelylow for the corresponding predicted CD34+ doses.

ABBREVIATIONS: AY = actual yield; CAS = Clinica Alemana

Santiago; China = The Affiliated Hospital of Soochow University;

HIAE = Hospital Israelita Albert Einstein; MDACC = The

University of Texas MD Anderson Cancer Center;

PB = peripheral blood; PBSC = peripheral blood stem cell;

PPV = positive predictive value; PY = predicted yield;

SCSP = Santa Casa Medical School.

From the 1Department of Stem Cell Transplantation and

Cellular Therapy, The University of Texas MD Anderson Cancer

Center, Houston, Texas; the 2Department of Hematology and

Department of Hemotherapy and Cellular Therapy, Hospital

Israelita Albert Einstein, and the 3Department of Hematology

and Oncology, Santa Casa Medical School, Sao Paulo, Brazil; the4Department of Hematology, The First Affiliated Hospital of

Soochow University, Jiangsu Institute of Hematology, Key

Laboratory of Thrombosis and Hemostasis, Ministry of Health,

Suzhou, P.R. China; the 5Department of Hematology, Clinica

Alemana, Santiago, Chile; and the 6Department of Medicine and

the Case Comprehensive Cancer Center, University Hospitals

Case Medical Center, Case Western Reserve University,

Cleveland, Ohio.

This research was supported in part by the National

Natural Science Foundation of China No. 81270645 and No.

81270617) and the Priority Academic Program Development of

Jiangsu Higher Education Institutions.

Presented in part at the American Society of Hematology

Meeting 2011.

Address reprint requests to: Chitra Hosing, MD, The

University of Texas MD Anderson Cancer Center, 1515

Holcombe Boulevard, Houston, TX 77030; e-mail:

[email protected].

Received for publication April 4, 2013; revision received

July 18, 2013, and accepted July 22, 2013.

doi:10.1111/trf.12435

TRANSFUSION 2014;54:1081-1087.

Volume 54, April 2014 TRANSFUSION 1081

Autologous or allogeneic stem cell transplants areroutinely performed for the management ofhigh-risk hematologic malignancies.1-5 In theautologous setting, stem cell collection is

achieved by mobilization of these cells into the peripheralblood (PB) with growth factors alone or preceding chemo-therapy.6 Mobilization in allogeneic donors is usuallyachieved with growth factors alone.7 Once adequatenumbers of stem cells are detected in the PB,leukapheresis is performed daily until a “target cell dose”is collected. Most transplantation physicians agree that aCD34+ dose of 2 × 106 cells/kg of recipient body weightis sufficient to proceed with high-dose chemotherapyand stem cell transplantation. Patients who proceed totransplantation with a suboptimal stem cell dose mayhave a worse outcome than patients who receive higherdoses.8,9

PB CD34+ counts are the strongest predictor of stemcell yield.10 In turn, stem cell mobilization into PB isdependent on variables such as donor age and sex, themobilization regimen, the diagnosis, and the number andtype of prior chemotherapy regimens.11-13 Although vari-able from institution to institution, most transplantcenters consider a minimum PB CD34+ count of10 × 106/L to be adequate for proceeding with stem cellcollection. The successful collection of stem cells dependson the collection efficiency (i.e., the ratio of the absolutenumber of stem cells collected via leukapheresis to thenumber of stem cells that pass through the cell separatorduring each procedure, multiplied by 100) and the bloodvolume processed.

Several attempts have been made to develop a math-ematical formula that would accurately predict the stemcell yield on the basis of preapheresis PB CD34+ counts,collection efficiency, donor and/or recipient weight, andblood volume processed.14-18 However, there is no univer-sal formula that has been shown to accurately predict thestem cell yield across different institutions. Pierelli andcolleagues16 proposed a mathematical formula to esti-mate the CD34+ dose collected on the first day ofapheresis on the basis of the preapheresis number of PBCD34+ cells, the amount of blood processed, and thedonor’s weight. Using the simplified formula, Pierelli andcoworkers16 found a high level of correlation between themeasured collection yields (or actual yield [AY]) and pre-dicted yield (PY; R = 0.85; p ≤ 0.0001) and a high degree ofprediction accuracy (AY vs. PY, p = 0.114781 by paired ttest; median AY-to-PY ratio, 1.23) in a cohort of 208patients who had undergone a total of 313 leukapheresisprocedures at seven institutions. They also found that theprediction accuracy of the formula was lower whenleukapheresis was performed with PB CD34+ counts of<15 × 106 cells/L. The formula also accurately predictedthe collection yields in normal donors whose stem cellswere mobilized with growth factors.

The aim of this study was to assess the predictivevalue of the simplified Pierelli formula in a larger cohort ofdonors across different institutions. Accurate prediction ofstem cell yield is important for collection centers in plan-ning leukapheresis sessions.

MATERIALS AND METHODS

Data were retrospectively collected on all consecutiveperipheral blood stem cell (PBSC) harvests conducted atfive institutions. Each institution selected a time frame forreported data. The five institutions were The University ofTexas MD Anderson Cancer Center in Houston, Texas(MDACC); Clinica Alemana in Santiago, Chile (CAS); Hos-pital Israelita Albert Einstein in Sao Paulo, Brazil (HIAE);The Affiliated Hospital of Soochow University, JiangsuInstitute of Hematology, Suzhou, P.R. China (China); andSanta Casa Medical School in Sao Paulo, Brazil (SCSP). Thelocal institutional review boards approved the retrospec-tive chart review protocol. Information on donor age, sex,diagnosis, weight, absolute peripheral CD34+ countswithin 24 hours of the first apheresis procedure, totalblood volume processed, and number of CD34+ cells har-vested per kilogram of recipient body weight was col-lected. Subjects for whom PB CD34+ counts were notavailable were not included in this analysis.

MDACCA computer-generated random 50% sample of the 851consecutive autologous PBSC harvests performedbetween January 1, 2005, and December 31, 2009, wasincluded in this analysis. The remaining 50% was pre-served for future validation studies. Allogeneic donorswere excluded because PB stem cell counts had not beenroutinely monitored at this institution before harvest.Thus, 397 subjects were included in this analysis. PBSCshad been collected after mobilization with chemotherapyplus growth factors or with high-dose growth factorsalone. The starting dose of filgrastim was 10 μg/kg/day(rounded to nearest vial size). Leukapheresis was startedwhen the PB CD34+ count had reached at least 10 × 106

cells/L. Three times the estimated blood volume was pro-cessed during each collection. ACD-A was used as antico-agulant. Samples from leukapheresis products had beencollected to determine the number of CD34+ cells beforetheir being cryopreserved. Yield of collected stem cells wascalculated per kilogram of recipient body weight. Alldonors had had central venous catheters placed for thecollection.

ChinaA total of 157 autologous PBSC harvests had been per-formed between January 1, 2001, and December 31, 2011.

HOSING ET AL.

1082 TRANSFUSION Volume 54, April 2014

PBSCs had been collected after mobilization with chemo-therapy plus growth factors or with high-dose growthfactors alone. The starting dose of filgrastim was 300 μg/day. Leukapheresis was started when the PB CD34+ counthad reached at least 10 × 106 cells/L. The anticoagulantused was ACD-A. The blood volume processed was basedon the weight of the patient and varied between 10 and12 L. A quarter of the donors had a central venous catheterplaced for the collection procedure.

CASA total of 186 donors who had had PBSCs collectedbetween January 1, 1995, and December 31, 2010, wereincluded in this analysis. All but three had been mobilizedwith chemotherapy followed by filgrastim (5 μg/kg onceor twice daily). All allogeneic donors had been mobilizedwith 5 μg/kg filgrastim twice daily. Leukapheresis wasstarted when PB CD34+ counts had been estimated to beat their highest (4 days after filgrastim had been started forallogeneic donors) and when white blood cell (WBC)counts had reached at least 4 × 109/L with PB monocytosisin autologous donors who had received chemotherapyand filgrastim. The total blood volume processed wasbetween 0.9 and 6.3 times the total individual bloodvolume. All autologous donors had had a central venouscatheter placed for leukapheresis.

HIAEA total of 297 patients had undergone 305 mobilizationattempts between February 1, 1999, and June 30, 2010.PBSCs had been collected with chemotherapy plus growthfactors or with high-dose growth factors alone. The start-ing dose of filgrastim was 10 μg/kg/day. PBSC collectionstarted when the PB CD34+ count was greater than10 × 106 cells/L. Large-volume leukapheresis (processingthree or four times the total blood volume) was employedto maximize the CD34+ yield per procedure. A centralvenous catheter had been inserted if a patient’s peripheralvenous access had been judged to be insufficient to main-tain an adequate inlet flow rate during apheresis.

SCSPA total of 83 apheresis procedures were performed in 2000.PBSCs had been collected after mobilization with 5 to10 μg/kg/day filgrastim or with chemotherapy followed byfilgrastim. Leukapheresis was initiated when the PBCD34+ count had reached at least 8 × 106 cells/L.

All sitesThe PB CD34+ cell counts were monitored by flowcytometry, and CD34+ cells had been gated, analyzed, and

reported as a percentage of WBCs using the InternationalSociety of Hematotherapy and Graft Engineering(ISHAGE) double platform.19 Only those collections inwhich a continuous inlet flow rate of at least 40 mL/minhad been obtained were included in this analysis.Leukapheresis had been performed daily to reach an indi-vidual targeted yield of at least 2 × 106 CD34+ cells/kg.All leukapheresis procedures were performed with acontinuous-flow blood cell separator (Cobe Spectra,CaridianBCT, Lakewood, CO) and a mononuclear cellor AutoPBSC protocol. The collected products werecryopreserved in dimethyl sulfoxide and stored in liquidnitrogen or in a mechanical freezer below −130°C.

Statistical analysisThe primary objective of the study was to validatePierelli’s simplified formula:16

Number of CD cells collected per kilogramof donor body

34+wweight Number of CD cells

per milliliter of PB mil= +(

) × ×34

0 4. lliliters of PBprocessed per kilogram of donor body weigh

(tt).

The formula estimates the average component col-lection efficiency of 40%.16 We assessed the relationshipbetween the PY calculated using the formula and the AY ofCD34+ cells that was collected on Day 1 of leukapheresis.To facilitate comparison of our findings with those previ-ously reported by Pierelli and colleagues, we first assessedthe linear relationship between AY and PY using Pearson’scorrelation coefficient and calculated the ratio of theyields (AY:PY). In addition, to characterize the relationshipbetween the yields in a clinically relevant manner, we clas-sified them as within the conventionally acceptable col-lected CD34+ doses (>2 × 106-5 × 106 cells/kg), below thisrange (≤2 × 106 cells/kg), or above it (>5 × 106 cells/kg) andassessed the likelihood that the Pierelli formula wouldcorrectly predict the range within which the AY would orwould not fall. Specifically, we estimated the positive pre-dictive value (PPV) to measure the likelihood that the AYwould correctly fall within the predicted range. We alsoestimated the 95% exact binomial confidence interval (CI)around the PPV. The primary interest of this study wasthe PPV. Higher values reflect a higher likelihood that thePierelli formula correctly identifies the range of values theAY would fall into. To determine whether the formula wasapplicable to both the adult and the pediatric populations,donors who were less than 18 years of age were analyzedseparately. Statistical analysis was performed using com-puter software (STATA 11.0, StataCorp, College Station,TX).

RESULTS

A total of 1126 leukapheresis procedures from five institu-tions were included in this analysis. The median age at the

PREDICTION OF STEM CELL YIELD


time of leukapheresis was 48 years (range, 2-80 years); 98donors (9%) were less than 18 years of age (Table 1). Forty-one percent of donors were female and 59% were male. Atall institutions, lymphoma and multiple myeloma werethe most common diagnoses. Allogeneic donors consti-tuted 24% of the CAS cohort (4% of the overall cohort).

Donors aged at least 18 yearsFor donors at least 18 years, the median PB CD34+ countwithin 24 hours of the first collection was 30 × 106/L(range, 0.3 × 106-2735 × 106/L). The median donor weightwas 74 kg (range, 34-164 kg). The median blood volumeprocessed was 16 L (range, 3-33 L). Thirty-two percent ofdonors collected more than 5 × 106 CD34+ cells/kg on thefirst day. The median number of CD34+ cells collected was3 × 106 cells/kg (range, 0.01 × 106-109 × 106 cells/kg). Themedian overall ratio of AY to PY was 1.1 (range, 0.02-27).This ratio was highest for the China and CAS cohorts(median of 1.5 for each). Pearson’s correlation coefficientbetween AY and PY was 0.76 overall (Table 2).

At 85%, PPV was highest for PY of more than 5 × 106

cells/kg (Table 3). Only 2% of cases predicted to have aCD34+ cell dose of more than 5 × 106 cells/kg had an AY ofnot more than 2 × 106 cells/kg. The lowest calculated PPVwas for PY more than 2 × 106 to 5 × 106 cells/kg (56%). Ofthese cases, 14% had an AY lower than the PY, while AY washigher than PY for 30%. The PPV for PY of not more than2 × 106/kg was 72%; thus, 28% of cases had an AY higherthan the PY. This pattern was consistent across institu-tions. PPV ranged from 65% (SCSP) to 79% (China) for PYof not more than 2 × 106/kg; from 42% (CAS) to 62%(MDACC) for PY of more than 2 × 106 to 5 × 106/kg; andfrom 71% (SCSP) to 93% (CAS and China) for PY of morethan 5 × 106/kg (exceeding 80% at all institutions exceptSCSP).

The results were equivalent when allogeneic donorswere excluded from the analysis (data not shown). Simi-larly, no significant differences were observed in subsetanalyses comparing lymphoma and multiple myelomapatients, the two largest diagnostic groups. The ability toaccurately predict the stem cell yield on the basis ofpreapheresis CD34+ cell counts was consistent acrossinstitutions. In most cases the AY exceeded the PY. Theonly group for which AY was underestimated was patientswho had had circulating CD34+ counts of 16 × 106 to30 × 106 cells/L. For example, of the 67 patients who werepredicted to have more than 2 × 106 to 5 × 106 CD34+cells/kg collected, 30% actually had not more than 2 × 106

cells/kg collected.

Donors aged less than 18 yearsDonors from MDACC were not included in this analysis asonly two patients were less than 18 years old. For the pedi-

TAB

LE

1.B

asel

ine

coh

ort

char

acte

rist

ics*

Cha

ract

eris

ticO

vera

ll(n

=11

26)

SC

SP

(n=

83)

CA

S(n

=18

6)C

hina

(n=

157)

HIA

E(n

=30

3)†

MD

AC

C(n

=39

7)

Year

ofda

taco

llect

ion

1995

-201

120

0019

95-2

010

2001

-201

119

99-2

010

2005

-200

9A

ge(y

ears

)48

(2-8

0)41

(10-

59)

35(2

-69)

37(8

-74)

48(4

-74)

56(1

7-80

)A

ge<

18ye

ars

98(9

)8

(10)

54(2

9)17

(11)

17(6

)2

(1)

Sex F

emal

e46

0(4

1)38

(46)

86(4

6)56

(36)

126

(42)

154

(39)

Mal

e66

6(5

9)45

(54)

100

(54)

101

(64)

177

58%

243

(61)

Dia

gnos

isA

cute

leuk

emia

(mye

loge

nous

orly

mph

obla

stic

)11

2(1

0)1

(1)

13(7

)62

(39)

36(1

2)0

(0)

Lym

phom

a(n

on-H

odgk

in’s

orH

odgk

in’s

)45

2(4

0)38

(46)

68(3

7)67

(43)

108

(36)

171

(43)

Mul

tiple

mye

lom

a41

3(3

7)44

(53)

38(2

0)26

(17)

79(2

6)22

6(5

7)C

hron

icly

mph

ocyt

icle

ukem

ia11

(1)

0(0

)0

(0)

0(0

)11

(4)

0(0

)C

hron

icm

yelo

geno

usle

ukem

ia5

(0)

0(0

)0

(0)

0(0

)5

(2)

0(0

)S

olid

tum

or54

(5)

0(0

)23

(12)

2(1

)29

(10)

0(0

)A

utoi

mm

une

dise

ase

35(3

)0

(0)

0(0

)0

(0)

35(1

2)0

(0)

Allo

gene

icdo

nor

44(4

)0

(0)

44(2

4)0

(0)

0(0

)0

(0)

*D

ata

are

repo

rted

asm

edia

n(r

ange

)or

num

ber

(%).

†A

tota

lof

297

patie

nts

unde

rwen

t30

3m

obili

zatio

nat

tem

pts.

HOSING ET AL.


atric donors in the CAS, China, HIAE, and SCSP cohorts(Table 4), the median PB CD34+ count within 24 hours ofthe first collection was 20 × 106 cells/L (range, 2 × 106-280 × 106 cells/L). The median donor weight was 39 kg(range, 9-89 kg). The median blood volume processed was10 L (range, 3-22 L). For 36% of patients, more than 5 × 106

CD34+ cells/kg had been collected on the first day. Themedian number of CD34+ cells collected was 3.6 × 106

cells/kg (range, 0.11 × 106-40 × 106 cells/kg). The medianoverall ratio of AY to PY was 1.4 (range, 0.1-17). Pearson’scorrelation coefficient between AY and PY was 0.73overall.

Overall, among donors aged less than 18 years, PPVwas highest for PY of more than 5 × 106 cells/kg (94%).None of the cases predicted to have a CD34+ count ofmore than 5 × 106 cells /kg had an AY of not more than2 × 106 cells/kg. As in the adult population, the PPV waslowest for PY of more than 2 × 106 to 5 × 106 cells/kg (31%).However, only 19% of these cases had an AY of not morethan 2 × 106 cells/kg, while 50% of cases had an AY higherthan the PY.

DISCUSSION

The purpose of this study was to validate across multipleinstitutions Pierelli’s predictive formula for CD34+ cellyield on leukapheresis. Our results suggest that thepreapheresis PB CD34+ count is associated with a highPPV for predicted CD34+ doses of more than 5 × 106

cells/kg and that the lowest PPV occurs with doses of morethan 2 × 106 to 5 × 106 cells/kg. These findings were con-sistent across institutions and were comparable for adultpatients and pediatric patients. In the majority of casesthe AY of stem cells was higher than the PY except inpatients who had had preapheresis PB CD34+ counts of16 × 106 to 30 × 106 cells/L. In this group, 30% of thosepredicted to have had a cell dose of more than 2 × 106 to5 × 106 CD34+ cells/kg collected had failed to collect theminimum acceptable dose of 2 × 106 cells/kg. In practice,collecting a higher stem cell dose than predicted has lessclinical impact than not meeting the predicted cell dose(in which case further collections would have to be sched-uled). The ability to accurately predict the stem cell yieldwas weakest for preapheresis CD34+ counts of 16 × 106 to30 × 106 cells/L. Thus, each treatment center could usepreapheresis PB CD34+ counts and the simplified Pierelliformula in predicting the stem cell yield and in planningleukapheresis procedures for their donors.

In a study by Trickett and colleagues,15 a retrospectiveanalysis of 67 leukapheresis procedures performed on 42patients, the CD34 prediction score correlated signifi-cantly with the harvested number of CD34+ cells/kg ofbody weight (r2 = 0.88, p < 0.0001). They prospectivelyconfirmed the validity of the relationship between theprediction score and the harvest number of CD34/kg by

TAB

LE

2.L

euka

ph

eres

isre

sult

sfo

rd

on

ors

atle

ast

18ye

ars

of

age*

Cha

ract

eris

ticO

vera

ll(N

=10

28)

SC

SP

(n=

75)

CA

S(n

=13

2)C

hina

(n=

140)

HIA

E(n

=28

6)M

DA

CC

(n=

395)

Abs

olut

eC

D34

+co

unt

(×10

6 /L)

30(0

.3-2

735)

31(4

.2-2

79)

32(1

-175

0)16

(0.3

-411

)31

(0.6

-604

)30

(2-2

735)

Don

orw

eigh

t(kg

)74

(34-

164)

69(4

0-14

3)70

(34-

110)

64(3

9-10

0)74

(40-

143)

84(4

5-16

4)D

ay1

bloo

dvo

lum

epr

oces

sed

(L)

16(3

-33)

14(1

0-25

)13

(3-2

9)10

(7-1

2)19

(8-3

3)16

(8-2

7)A

ctua

lDay

1C

D34

+do

se(c

ells

×106 /

kg)

3(0

.01-

109)

3(0

.2-2

6)4

(0.0

1-10

9)2

(0.0

2-50

)3.

5(0

.1-4

4)3

(0.2

-101

)≤2

383

(37)

25(3

3)48

(36)

73(5

2)98

(34)

139

(35)

>2-5

313

(30)

26(3

5)33

(25)

29(2

1)77

(27)

148

(37)

>533

2(3

2)24

(32)

51(3

9)38

(27)

111

(39)

108

(27)

Pre

dict

edD

ay1

CD

34+

dose

†(c

ells

×106 /

kg)

2.4

(0.0

2-21

0)2.

6(0

.3-2

4)2.

3(0

.1-7

3)0.

98(0

.02-

22)

3.2

(0.1

-67)

2.4

(0.1

-210

)≤2

468

(46)

34(4

5)62

(47)

90(6

4)10

3(3

6)17

9(4

5)>2

-528

6(2

8)17

(23)

40(3

0)22

(16)

77(2

7)13

0(3

3)>5

274

(27)

24(3

2)30

(23)

28(2

0)10

6(3

7)86

(22)

Rat

ioof

pred

icte

dto

actu

alC

D34

+do

se1.

1(0

.02-

27)

1.1

(0.2

-11)

1.5

(0.0

3-6)

1.5

(0.0

2-27

)1.

0(0

.04-

12)

1.1

(0.2

-10)

Pea

rson

’sco

rrel

atio

nco

effic

ient

0.76

0.84

0.86

0.85

0.79

0.79

*D

ata

are

repo

rted

asm

edia

n(r

ange

)or

num

ber

(%).

†P

iere

lli’s

sim

plifi

edfo

rmul

a:C

D34

+pr

edic

tion

scor

e=

(num

ber

ofC

D34

+ce

llspe

rm

illili

ter

ofP

B)

×0.

4×

(mill

ilite

rsof

PB

proc

esse

dpe

rki

logr

amof

dono

rbo

dyw

eigh

t).



analyzing data from 24 leukapheresis procedures per-formed on 20 subsequent patients. For example, based ontheir data for a target cell dose of at least 2 × 106 CD34+cells/kg, a prediction score of more than 6.5 would need tobe attained. All leukapheresis procedures with a predic-tion score of more than 6.5 achieved the target number ofat least 2 × 106 CD34+ cells/kg in a single procedure. Thestrong correlation between the CD34 prediction score andthe number of harvested CD34+ cells/kg led the authors toreconfigure the algorithm to determine the minimumblood volume required to be processed by leukapheresisto yield the target number of CD34+ cells/kg.

One of the strengths of our study is the validation ofthe simplified Pierelli formula in a large multicenter

cohort of donors. Moreover, the results were applicableto both adult and pediatric population. As expected,for preleukapheresis PB CD34+ counts of more than60 × 106/L the PPV was high. The PPV was the weakest indonors who had preapheresis PB CD34+ counts between16 × 106 and 30 × 106 cells/L. A further analysis of factorsthat may more accurately predict the stem cell yield in thissubgroup is warranted. The major limitation is the retro-spective nature of the study.

In conclusion, the formula developed by Pierelli andcolleagues is associated with a PPV that is high, moderate,and relatively low for predicted CD34+ doses of more than5 × 106, not more than 2 × 106, and more than 2 × 106 to5 × 106 cells/kg, respectively. Except for patients with

TABLE 3. Actual versus predicted CD34+ yields in donors at least 18 years of age

Predicted CD34+ yield

Actual CD34+ yield (%)

PPV (%) 95% CI≤2 × 106 cells/kg >2 × 106-5 × 106 cells/kg >5 × 106 cells/kg

Overall≤2 × 106 cells/kg 72 25 3 72 68-76>2-5 × 106 cells/kg 14 56 30 56 50-62>5 × 106 cells/kg 2 14 85 85 80-89

SCSP≤2 × 106 cells/kg 65 29 6 65 46-80>2-5 × 106 cells/kg 12 59 29 59 33-81>5 × 106 cells/kg 4 25 71 71 49-87

CAS≤2 × 106 cells/kg 74 22 3 74 61-84>2-5 × 106 cells/kg 5 42 52 42 27-59>5 × 106 cells/kg 0 7 93 93 78-99

China≤2 × 106 cells/kg 79 17 4 79 69-87>2-5 × 106 cells/kg 4 59 36 59 36-79>5 × 106 cells/kg 4 4 93 93 76-99

HIAE≤2 × 106 cells/kg 75 24 1 75 65-83>2-5 × 106 cells/kg 25 49 26 49 38-61>5 × 106 cells/kg 2 13 85 85 77-91

MDACC≤2 × 106 cells/kg 69 29 2 69 61-75>2-5 × 106 cells/kg 12 62 25 62 53-71>5 × 106 cells/kg 0 17 83 83 73-90

TABLE 4. Leukapheresis results in patients less than 18 years of age*Characteristic Overall (N = 98) SCSP (n = 8) CAS (n = 54) China (n = 17) HIAE (n = 17)

Absolute CD34 (×106/L) 20 (2-280) 36 (15-180) 15 (5-119) 10 (2-280) 29 (6-277)Weight (kg) 39 (9-89) 61 (38-78) 28 (9-72) 59 (22-89) 35 (13-83)Blood volume processed on Day 1 (L) 10 (3-22) 13 (8-16) 8.5 (3-22.5) 9.5 (4.5-12) 11.5 (4.6-19)Actual Day 1 CD34+ dose (cells ×106/kg) 3.6 (0.11-40) 3.8 (1-29) 3.9 (0.5-14) 1.5 (0.11-40) 2.2 (0.2-33)

≤ 2 35 (36) 2 (25) 15 (28) 9 (53) 8 (47)>2-5 28 (29) 2 (25) 21 (39) 3 (18) 1 (6)>5 35 (36) 4 (50) 18 (33) 5 (29) 8 (47)

Predicted Day 1 CD34+ dose† (cells ×106/kg)≤2 46 (47) 4 (50) 28 (52) 10 (59) 4 (24)>2-5 36 (37) 2 (25) 24 (44) 4 (24) 5 (29)>5 16 (16) 2 (25) 2 (4) 3 (18) 8 (47)

Ratio CD34+ yield actual/predicted 1.4 (0.1-17) 1.10 (0.8-2) 1.5 (0.6-6) 1.7 (0.5-17) 0.9 (0.1-1.8)Pairwise correlation coefficient 0.73 0.90 0.81 0.90 0.62

* Data are reported as median (range) or number (%). MDACC not included in this table as only two patients were less than 18 years old.† Pierelli’s simplified formula: CD34+ prediction score = (number of CD34+ cells per milliliter of PB) × 0.4 × (milliliters of PB processed per

kilogram of donor body weight).

HOSING ET AL.


preapheresis CD34+ counts of 16 × 106 to 30 × 106 cells/L,the AY was equal to or higher than the PY.

CONFLICT OF INTEREST

The authors report no conflicts of interest.

REFERENCES

1. Rajkumar SV. Multiple myeloma: 2012 update on diagno-

sis, risk-stratification, and management. Am J Hematol

2012;87:78-88.

2. Harousseau JL, Moreau P. Autologous hematopoietic

stem-cell transplantation for multiple myeloma. N Engl J

Med 2009;360:2645-54.

3. Nieto Y, Popat U, Anderlini P, et al. Autologous stem-cell

transplantation for refractory or poor-risk relapsed Hodg-

kin’s lymphoma: effect of the specific high-dose chemo-

therapy regimen on outcome. Biol Blood Marrow

Transplant 2013;19:410-7.

4. Blanes M, de la Rubia J. Role of autologous bone marrow

transplant in multiple myeloma. Curr Opin Oncol 2012;

24:733-41.

5. van Kampen RJ, Canals C, Schouten HC, et al. Allogeneic

stem-cell transplantation as salvage therapy for patients

with diffuse large B-cell non-Hodgkin’s lymphoma relaps-

ing after an autologous stem-cell transplantation: an

analysis of the European Group for Blood and Marrow

Transplantation Registry. J Clin Oncol 2011;29:1342-8.

6. Pesek G, Cottler-Fox M. Hematopoietic stem cell

mobilization: a clinical protocol. Methods Mol Biol 2012;

904:69-77.

7. Anderlini P, Przepiorka D, Seong C, et al. Factors affecting

mobilization of CD34+ cells in normal donors treated

with filgrastim. Transfusion 1997;37:507-12.

8. Gertz MA, Wolf RC, Micallef IN, et al. Clinical impact and

resource utilization after stem cell mobilization failure in

patients with multiple myeloma and lymphoma. Bone

Marrow Transplant 2010;45:1396-403.

9. Pavone V, Gaudio F, Console G, et al. Poor mobilization is

an independent prognostic factor in patients with malig-

nant lymphomas treated by peripheral blood stem cell

transplantation. Bone Marrow Transplant 2006;37:719-24.

10. Ford CD, Pace N, Lehman C. Factors affecting the effi-

ciency of collection of CD34-positive peripheral blood

cells by a blood cell separator. Transfusion 1998;38:1046-

50.

11. Popat U, Saliba R, Thandi R, et al. Impairment of

filgrastim-induced stem cell mobilization after prior

lenalidomide in patients with multiple myeloma. Biol

Blood Marrow Transplant 2009;15:718-23.

12. Hosing C, Saliba RM, Ahlawat S, et al. Poor hematopoietic

stem cell mobilizers: a single institution study of inci-

dence and risk factors in patients with recurrent or

relapsed lymphoma. Am J Hematol 2009;84:335-7.

13. Vasu S, Leitman SF, Tisdale JF, et al. Donor demographic

and laboratory predictors of allogeneic peripheral blood

stem cell mobilization in an ethnically diverse popula-

tion. Blood 2008;112:2092-100.

14. Humpe A, Riggert J, Munzel U, et al. A prospective, ran-

domized, sequential, crossover trial of large-volume

versus normal-volume leukapheresis procedures: effect

on progenitor cells and engraftment. Transfusion 1999;39:

1120-7.

15. Trickett AE, Smith S, Kwan YL. Accurate calculation of

blood volume to be processed by apheresis to achieve

target CD34+ cell numbers for PBPC transplantation.

Cytotherapy 2001;3:5-10.

16. Pierelli L, Maresca M, Piccirillo N, et al. Accurate predic-

tion of autologous stem cell apheresis yields using a

double variable-dependent method assures systematic

efficiency control of continuous flow collection proce-

dures. Vox Sang 2006;91:126-34.

17. Cottler-Fox MH, Lapidot T, Petit I, et al. Stem cell mobili-

zation. Hematology Am Soc Hematol Educ Program 2003;

419-37.

18. Rosenbaum ER, O’Connell B, Cottler-Fox M. Validation of

a formula for predicting daily CD34(+) cell collection by

leukapheresis. Cytotherapy 2012;14:461-6.

19. Sutherland DR, Anderson L, Keeney M, et al. The ISHAGE

guidelines for CD34+ cell determination by flow

cytometry. International Society of Hematotherapy and

Graft Engineering. J Hematother 1996;5:213-26.



peripheral blood stem cell yield calculated using preapheresis absolute cd34+ cell count, peripheral...

Documents