www.elsevier.com/locate/transci

Transfusion and Apheresis Science 30 (2004) 23–28

Blood component fractionation: manual versusautomatic procedures

Daniela Pasqualetti a,*, Alessandro Ghirardini b, Maria Cristina Arista a,Stefania Vaglio a, Azis Fakeri a, Alan A. Waldman c, Gabriella Girelli a

a Department of Cell Biotechnology and Hematology, University ‘‘La Sapienza’’, Blood Bank, Via Chieti 7, Rome 00161, Italyb Istituto Superiore di Sanit�a, Laboratory of Epidemiology and Biostatistics, Rome, Italy

c Waldman Biomedical Consultancy, Oceaside, New York, USA

Abstract

Over the last few years, quality system requirements have been introduced for blood components. The necessary

compliance with standard productions will have a considerable impact on Blood Banks. The introduction of automated

methods is the most satisfactory means to meet these requirements for blood component preparation.

A new device has been developed to automate the fractionation of blood into components. We evaluated the efficacy

of this instrument as compared to manual methods. A total of 218 units of blood have been collected, into several

different commercial blood bag systems (77 into standard quadruple bag systems, 141 into bag systems with integrated

in line filters), and used to evaluate the universality of the instrument. Whole blood units were processed using the Top/

Top system and the Compomat G4 (Fresenius HemoCare). A separate program protocol was developed for each kind

of bag.

Use of the Compomat G4 resulted in a statistically significant (p < 0:001) increase of the hemoglobin in filtered red

cell concentrates (RCC) in comparison with the manual procedure, and a similar trend, even not statistically significant,

has been observed for filtered RCC. Regardless of bag systems, we were able to observe a statistically significant in-

crease of platelets in the platelet concentrates (PCs), when comparing automatic versus manual procedure.

The automated procedure has been shown to be fast, and easy for the operators. This device reliably produces

acceptable blood components, and has been shown adaptable to use with different blood bag systems.

� 2003 Elsevier Ltd. All rights reserved.

1. Introduction

The implementation of a quality system for

blood component production has been much dis-

* Corresponding author. Tel.: +39-06-85795-524; fax: +39-

06-85795-501.

E-mail address: pasqualetti@bce.uniroma1.it (D. Pasqua-

letti).

1473-0502/$ - see front matter � 2003 Elsevier Ltd. All rights reserve

doi:10.1016/j.transci.2003.07.002

cussed over the last few years, and bringing stan-

dard production into compliance will have a

considerable impact on Blood Banks. In actual

practice, it often has been necessary to change

working methods in order to guarantee repro-

ducible production of the blood components,

introducing careful controls of the processing

procedures involved and performing quality testsof the blood products on a routine basis. While in

the past the procedures for preparing fractions of

d.

24 D. Pasqualetti et al. / Transfusion and Apheresis Science 30 (2004) 23–28

blood components were based completely on

manual operations, automated methods have re-

cently been introduced as the most efficient and

effective way to fulfil the standards for bloodcomponents production and conform with the

European guidelines [1].

Using collection bags which allow for removal

of materials only through a port on the top (Top/

Top system), standard blood fractionation in the

United States begins with the separation, by low-

speed centrifugation, of platelet-rich plasma

(PRP), red cell concentrates (RCC) and buffy-coat(BC) from whole blood (WB). In a subsequent

high-speed centrifugation, platelet concentrates

(PCs) and platelet poor plasma (PPP) are derived

from the PRP [2]. In Europe, in the 1990�s, a dif-

ferent approach was introduced, one designed to

decrease development of the platelet storage lesion

and in vitro activation due to the second centri-

fugation [3,4]. Using collection bags which allowfor removal of materials through ports on both the

top and the bottom (Top/Bottom system), it is

possible, following an initial high-speed centrifu-

gation of the WB, to produce PPP on the top, with

the platelets concentrating into a BC on a cushion

of RCC at the bottom [5]. After collection of this

BC, the platelets can be separated by a gentle low-

speed centrifugation, a process that presumablyproduces less damage than high-speed packing,

which causes formation of tight pellets of plate-

lets [6].

In actual practice, however, examination of

platelets stored for up to five days indicated no

significant differences in platelet survival and

function with the two preparative techniques [7,8],

and it is still being debated whether or not plateletsfrom PRP or from BC are better [8].

While there have been several studies performed

to validate the performance of an automatic

instrument with the Top & Bottom system [9–11],

the results of using an automatic extractor for the

Top/Top system for blood component preparation

have not been reported.

As part of our efforts to improve productquality and to reduce manufacturing problems, we

have tested the Compomat G4 (Fresenius Hemo-

Care) equipment developed to automate Top/Top

fractionation, and compared the quality of the

components produced with those obtained by the

standard manual method. As part of this study, we

evaluated Top/Top blood bag systems from dif-

ferent manufacturers.

2. Materials and methods

To check on the universality of the automatic

instrument Compomat G4 (Fresenius HemoCare),

we collected from blood donors 218 blood units,

each of 450 ml ± 10%, into 63 ml of CPD antico-

agulant solution, using blood bag collection sys-

tems from different manufacturers.Standard quadruple bag systems were used for

77 of the blood collections, while blood bag

collection systems with integrated in-line filters

were used for 141 of the blood collections. In the

first group of bag systems, manual procedures

were used to manufacture components from 13

units collected into Maco Pharma bags, 12 units

collected into Terumo bags, and 10 units col-lected into Fresenius bags, while the automatic

device was used to manufacture components

from 8 units collected into Maco Pharma bags,

20 units collected into Terumo bags, and 14

units collected into Fresenius bags. In the second

group of bags system with in-line filters, manual

procedures were used to manufacture compo-

nents from 80 units collected into Fresenius bagsand from 18 units collected into bags manufac-

tured by Baxter, while the automated device was

used to manufacture components from 31 units

collected into Fresenius bags and 12 units col-

lected into Baxter bags.

The quadruple bag systems were processed

immediately after blood collection, while all the in-

line bag systems were processed after 2 h sinceblood collection and after cooling at a conditioned

temperature between 18 and 22 �C. All the WB

units were centrifuged at 550g for 8 min at 20 �C,with an acceleration rate of 2 and deceleration 1 in

a Beckman J-6M/E centrifuge, to produce RCC.

PC were prepared from PRP by a second centri-

fugation at 2200g for 10 min.

The automated fractionation was performedusing Compomat G4 (Fresenius-HemoCare).

D. Pasqualetti et al. / Transfusion and Apheresis Science 30 (2004) 23–28 25

Following each of the centrifugations, the blood

packs are taken out of the bucket, the primary

bag hung on the pins on the front of the machine,

and the satellite bags placed on the top of theCompomat G4, with the integrated tubing being

placed in built-in automated clamper/sealers. After

the first centrifugation of the WB, as directed

by the software, three presses combined with

the integrated detectors express out PRP and,

if desired, BC, into satellite bags, leaving the

RCC. After the second centrifugation, the pre-

sses and integrated detectors express out the PPPinto a satellite bag, leaving the PCs. The detectors,

the sensitivity of which were set to meet our stan-

dards, allow control over the blood component

separation, and standardization for different bags.

The separation of the 141 blood units with in-

line filters was performed with the same procedure

as described above, placing the filter upside down

on the right hand side panel of the Compomat. Inaddition, this equipment has been used on Frese-

nius filters with the partial squeezing of Sag-M

through the filter by means of the top press. The

filtration of the 141 RCC with in-line filters was

carried out by gravity.

By using Compomaster, a software running on

Windows NT that manages the Compomat, the

Compomat G4 is able to process the blood unitswith a programmed and personalized protocol for

each kind of bag. The Compomaster/Compomat

programs every procedure, from the start to end of

separation, and registers all the data. At our site,

in order to obtain the best recovery results for the

RCC, defined as lowest hemoglobin loss into the

small BC volume and a low level of WBC con-

tamination, we developed specific parameters forthe processing programs for the different bag sys-

tems, as follows:

(a) Fresenius and Maco Pharma blood bag sys-

tems:

The upper press is pushed out to 54.0 mm and

the red cell level reaches first the DET A2 by

means of both presses, then the DET A6+B usingonly the lower press with a speed 20% to reduce

the squeezing flow. The combination of these

parameters mainly defines the features of the BC,

in terms of volume and hematocrit.

(b) Terumo and Baxter blood bag systems:

Upper press out to 53.0 mm, both presses out

DET A1 and lower press out DET A8+B with a

speed 25%.(c) Fresenius and Baxter blood bag systems with

in-line filters:

Upper press out 50.0 mm, lower press out DET

A6 with speed 15% and then lower press out DET

A8+B with speed 20%.

(d) PRP derived from all systems:

Both presses to 45.0 mm, upper press to 56.0

mm and lower press to 58.0 mm.

After standardization of the method for stain-

ing of leukocytes with a commercial kit (Leuco-

Count, Becton Dickinson, San Jose, Costa

Mesa-CA, USA), this kit and calibrated flow cyto-

metry (FACScan, Becton Dickinson, San Jose,

Costa Mesa-CA, USA) were used to count the

absolute number of the residual [12]. Otherhematological parameters were measured with an

automated analyser (AcT.5DiffCP-Coulter, Instru-

mentation Laboratory, Milan, Italy).

Statistical analysis was performed by comparing

mean value of the relevant variables, by using the

Epi Info software (version 6) produced by the

Centers forDisease Control andPrevention (CDC).

3. Results

The average gain in hemoglobin recovery in

RCC was equal to or more than 3 g/unit for both

filtered and not filtered RCC. As shown in Tables

1 and 2, the automatic procedure is associated with

a statistically (p < 0:001) significant increase in

hemoglobin recovery in filtered RCC, while for

RCC the difference does not reach the significant

level (p ¼ 0:12). The apparent greater increase inred cell recovery for the Fresenius bags (Table 3) is

felt, in part, to be due to the large standard devi-

ation in the volumes of the individual collections.

With regard to the higher values for hemoglo-

bin in the filtered RCC (Table 2), this is apparently

due to a decreased loss in the filters, a loss that, for

the Baxter bags, falls to 11.17% (Table 4).

It should be noted that the Fresenius filterwas hard-housing and the Baxter filter was

Table 1

Comparison of results between manual and automatic proce-

dure in standard bag systems from all different manufactures

Overall summary

of standard bag

systems

Manual procedure

(N ¼ 35)

Compomat G4

(N ¼ 42)

Mean (SD) Mean (SD)

RCC

Volume (ml) 335.7 (22.6) 351.6 (24.6)

Hb (g/unit) 61.0 (7.7) 64.6 (5.7)

Ht (%) 52.9 (5.3) 53.2 (2.2)

PCs

Plts (·10e11) 0.62 (0.17) 0.81 (0.2)

Recovery (%) 73.8 (14.1) 92.6 (14.5)

Table 2

Comparison of results between manual and automatic proce-

dure in in-line filter systems from all different manufactures

Overall Summary of

in-line filter systemsManual procedure

(N ¼ 98)

Compomat G4

(N ¼ 43)

Mean (SD) Mean (SD)

Filtered RCC

Hb (g/unit) 53.3 (7.8) 59.2 (7.8)

Hb loss (%) 13.0 (7.6) 10.4 (4.4)

WBC Log depletion 4.36 (0.6) 4.14 (0.3)

PCs

Plts (·10e11) 0.61 (0.2) 0.71 (0.3)

Recovery (%) 73.5 (15.1) 82.0 (24)

Table 3

Comparison of the results produced by Compomat G4 of red

cell concentrates using bag systems from three different manu-

facturers

Results with

different bag

system

manufacturers

Terumo

(N ¼ 20)

Fresenius

(N ¼ 14)

Maco Pharma

(N ¼ 8)

Mean (SD) Mean (SD) Mean (SD)

RCC

Volume (ml) 340.0 (12.4) 370.4 (25.1) 365.4 (18)

Hb (g/unit) 62.6 (3.2) 68.4 (6.5) 57.2 (3.9)

Ht (%) 52.8 (2.1) 53.8 (2.6) 52.8 (2.2)

PCs

Plts (·10e11) 0.82 (0.16) 0.85 (0.3) 0.60 (0.33)

Recovery (%) 92.9 (11.8) 95.0 (14.4) 67.1 (10.2)

Table 4

Comparison of the results produced by Compomat G4 of fil-

tered red cell concentrates with bag systems with in-line filters

from the two different manufacturers used

Results with the two

bag systems with in-line

filters manufacturers

Fresenius

(N ¼ 31)

Baxter

(N ¼ 12)

Mean (SD) Mean (SD)

Filtered RCC

Hb (g/unit) 58.1 (7.8) 62.09 (5.24)

Hb loss (%) 10.4 (4.4) 11.17 (8.6)

WBC Log depletion 4.14 (0.30) 4.08 (0.27)

PCs

Plts (·10e11) 0.70 (0.3) 0.68 (0.7)

Recovery (%) 84 (22) 76 (30)

26 D. Pasqualetti et al. / Transfusion and Apheresis Science 30 (2004) 23–28

soft-housing. The priming of the first, before thepassive filtration, was performed automatically by

use of the top press, while for the second both

priming and filtration was carried out by gravity.There was no statistically significant difference

(p ¼ 0:52) in residual WBC (Table 4) for both

types of WBC filtered components.

The results for PCs from the standard bags are

presented in Table 1. A statistically significant

difference (p < 0:003) has been observed between

the automatic procedure compared to manual

procedure, with a remarkable increment of theplatelet count, roughly 20%, in the PCs. Table 3

presents the results obtained for each of several

different manufacturers. The automatic prepara-

tion method was found to yield excellent and

comparable platelet counts with both the Terumo

and Fresenius blood bag systems, with a recovery

greater than 92.9% and 95.0%, respectively. This is

a statistically (p < 0:001) significant difference, ascompared to the results with the Maco Pharma

bags, which had a recovery of 62.1%.

Similarly, the automated method proved supe-

rior to the manual method for the preparation

of PCs from blood bag systems with filters, with

a mean recovery per unit of 0.71 · 1011 versus

0.61 · 1011 (Table 2). Table 4 presents the results

obtained for the two different bag manufacturers.No statistically significant differences were found

in the results obtained with the Fresenius and

Baxter blood bag systems with in-line filters.

The average processing time using the Compo-

mat was 2 min and 15 s for each bag. The total

time for the procedure using automatic equipment

was 4 min and 50 s, while the total time for the

manual operation was 6 min and 36 s.

D. Pasqualetti et al. / Transfusion and Apheresis Science 30 (2004) 23–28 27

There were no problems with the equipment,

and no breakdowns during the study. The only

unexpected events during the study were two

leaking seals, failures traced to incorrect posi-tioning of the tubing in the welding clamp.

4. Discussion

It is well known that, for transfusion therapy

to be effective, there must be quality and consis-

tent blood components. This study provides data

on the performance of an available automated

instrument, Compomat G4 (Fresenius Hemo-

Care), for separating blood components. Perfor-mance, the accuracy of the separation process, and

the reliability of the device using different bag

systems, were evaluated and compared with the

manual method previously used in our Blood

Center.

Performance, accuracy and reproducibility of

the automatic method were assessed in two ways,

both on ability to meet the standards in theEuropean guidelines, and on the extent of varia-

tion in the results. The results obtained for the

components produced were all well above the

standards of the European guidelines, and the

standard deviation of the parameters measured

were either the same or lower for the automatic

procedure than for the manual procedure.

Our results demonstrate that when this devicewas used to prepare red cell concentrates (RCC)

following a low-speed centrifugation to prepared

PRP, the level of hemoglobin in the RCC was

higher than for the manual method. When this

device was used in conjunction with in-line filters,

there was less loss of hemoglobin during prepara-

tion of the filtered RCC than for the manual

method. In Europe, it is required [1] that RCC andfiltered RCC units contain more than 43 and 40 g

hemoglobin/unit, respectively. This study suggests

that, in the light of the ongoing standardization of

these methods, higher values for the expected lev-

els of hemoglobin could be set.

Similarly, our results demonstrate that when

this device is used to separate PPP to yield PC,

there is a marked improvement in the plateletcontent of the component. Indeed, the level of

recovery was high enough to allow us to reduce the

number of units used for the preparation of pooled

PCs. This reduces the risk of exposure to infectious

agents and to allogeneic stimuli as well as reducingthe costs of this product. In our facility, where the

increased request for PCs does not allow storage of

the PCs for more than one or at most two days, we

have established that it is best to prepare platelets

by this two-step method, using the automated

device Compomat G4.

Usage of pools of PC made by the two-step

centrifugation method has, in the past, beenquestioned. Since the beginning of the 1990�s in the

United States, in order to reduce risks of alloim-

munization, there has been an increase in the use

of apheresis-derived platelet concentrates (APCs),

rather than single unit PCs, for treatment of

thrombocytopenic patients. In 1994, more than

half of all transfused platelets were performed with

APC�s [13], with markedly increased costs. How-ever, the TRAP Study group [14] demonstrated

that, when blood products are leucodepleted, there

is no advantage of APCs over PCs with regard to

alloimmunization rates. It also was suggested, by

several studies in the past, that the quality of BC-

PCs was better than PRP-PCs [15,16]. However,

recent studies demonstrate that there are no dif-

ferences in low morphological score and recoverybetween filtered pooled BC-PCs and PRP-PCs [17]

and that, in vitro, the differences of activation,

fibrinogen and glycoproteins were not significant

after one day of storage [7].

Our Center processes approximately 20,000

blood units/year, and for each product uses a

specific kind of bag, with the aim of improving the

quality process, so we evaluated the universalapplicability of the instrument, using the same

separation process with bag systems from several

different manufacturers. The results confirm that

this device can be used successfully with bags from

numerous sources.

In conclusion, given the improvement in the

production and content of the blood components

produced, the compliance with quality require-ments and the easy, rapid and reproducible tech-

nology, we would recommended the use of the

Compomat/Compomaster system to process blood

into its components.

28 D. Pasqualetti et al. / Transfusion and Apheresis Science 30 (2004) 23–28

Acknowledgements

The authors express their appreciation to Mrs

M.L. Salvagnini and Mrs G. Alimenti for thetechnical work and for help with the preparation

of the samples tested in this study. The authors are

also grateful to Dr. Chiara Franciosi for helpful

discussions.

References

[1] Guide to the preparation, use and quality assurance of

blood components. 6th ed. Strasbourg: Council of Europe

Publishing; 2000.

[2] The American Association of Blood Banks Annual Report

1990. Arlington, Va, American Association of Blood

Banks, 1990.

[3] Prins HK, de Bruijn JCGH, Henrichs HPJ, Loos JA.

Prevention of microaggregate formation by removal of

‘‘buffycoats’’. Vox Sang 1980;39:48–51.

[4] Fijnheer R, Veldman HA, van der Eertwegh AJ, Gouw-

erok CW, Homburg CH, Boomgaard MN, et al. In vitro

evaluation of buffy-coat-derived platelet concentrates

stored in a synthetic medium. Vox Sang 1991;60:16–22.

[5] Hogman CF, Eriksson L, Hedlund K, Wallvik J. The

bottom and top system: a new technique for blood

component preparation and storage. Vox Sang 1988;55:

211–7.

[6] Fijnheer R, Pieters RNI, de Korte D, Gouwerok CWN,

Dekker WJA, Reesink HW, et al. Platelet activation during

preparation of platelet concentrates: a comparison of the

platelet-rich plasma and the buffy coat methods. Transfu-

sion 1990;30:634–8.

[7] Lozano M, Estebanell E, Cid J, Diaz-Ricart M, Mazzara

R, Ordinas A, et al. Platelet concentrates prepared and

stored under currently optimal conditions: minor impact

on platelet adhesive and cohesive functions after storage.

Transfusion 1999;39:951–9.

[8] Anderson KC, Ness PM. Scientific basis of Transfusion

medicine. 2nd ed. Implications for clinical basis. Saunders

Company; 2000.

[9] Kretschmer V, Tesche U, Richter P. Improvement of the

separation of blood by modification of separators Opti-

press and Biotrans Separator. Beitr Infus 1992;30:112–21.

[10] Hurtado C, Bonanad S, Soler MA, Mirabet V, Blasco I,

Planelles MD, et al. Quality analysis of blood components

obtained by automated buffy-coat layer removal with a

Top & Bottom system (Optipress II). Haematologica

2000;85:390–5.

[11] van Rhenen DJ, Vermeij J, de Voogt J, Bernes JC, Payrat

JM. Quality and standardization in blood component

preparation with an automated blood processing tech-

nique. Transfus Med 1998;8:319–24.

[12] Dzik S, Moroff G, Dumont L, for the BEST Working

Party of the ISBT. A multicenter study evaluating three

methods for counting residual WBCs in WCB-reduced

blood components: nageotte hemocytometry, flow cytom-

etry, and microfluorometry. Transfusion 2000;40:513–20.

[13] Wallace EL, Churchill WH, Surgenor DM, Cho GS,

McGurk S. Collection and transfusion of blood and blood

components in the United States, 1994. Transfusion

1998;38:625–36.

[14] The Trial to reduce Alloimmunization to Platelet Study

Group. Leukocyte reduction and ultraviolet B irradiation

of platelets to prevent alloimmunization and refratoriness

to platelet transfusions. N Engl J Med 1997;25:1861–9.

[15] KretschmerV,BierannE, LohH. Separation of platelet con-

centrates (PC) from buffy-coats (BC) using the bottom and

top (BAT) drainage system. Transfus Sci 1990;11:363–6.

[16] Rebulla P, Bertolini F, Porretti L, Marangoni F, Smacchia

C, Marconi M, et al. Platelet concentrates from buffy coats:

improved conditions for preparation and evaluation in

routine clinical use. Transfus Sci 1993;14:41–6.

[17] Kostelijk EH, Klomp AJA, Engbers GHM, Gouwerok

CWN, Verhoeven AJ, van Aken WG, et al. Improved

platelet compatibility of water vapour glow discharge

treated non woven poly(ethylene terephthalate)leukocyte-

reduction filters for different types of platelet concentrates.

Transfus Med 2001;11:199–205.