Sterilization method of platelet storage containers affects in vitro parameters.

Abstract

Four butyryl trihexyl phthalate plasticized polyvinyl chloride (BTHC-PVC) containers were compared for storage of leukoreduced platelet concentrates (LR-PC): three ethylene oxide (EtO) sterilized (Gambro, Haemonetics, Fresenius), and one steam-sterilized (Fresenius).
LR-PCs were made from 5 buffy coats and 300 ml Composol additive solution, and leukoreduced by filtration. Four LR-PCs were pooled and subsequently divided over the 4 BTHC-PVC bags to prevent donor-dependent differences, and sampled for in vitro analysis on day 1, 2, 5, 7 and 9.
The pH values on day 9 were (mean +/- SD, n = 10): 7.12 +/- 0.03 (Gambro), 7.12 +/- 0.04 (Haemonetics), and 7.07 +/- 0.09 (Fresenius, EtO-sterilized) (not significantly different), vs. 6.91 +/- 0.12 (Fresenius, steam-sterilized; P < 0.001 vs. all EtO-sterilized bags). LR-PCs stored in the steam-sterilized bag exhibited significantly higher glucose consumption and lactate production (P < 0.001 vs. all EtO-sterilized bags).
All BTHC-PVC containers allow storage of LR-PCs for up to 9 days with good in vitro parameters. However, the method of sterilization affects the storage conditions of the LR-PCs in these bags.

Full text

Page 1

32
Vox Sanguinis
(2007)
92
, 32–36
ORIGINAL PAPER
©
2006 The Author(s)
Journal compilation
©
2006 Blackwell Publishing Ltd.
DOI: 10.1111/j.1423-0410.2006.00855.x
Blackwell Publishing Ltd
Sterilization method of platelet storage containers affects
in vitro
parameters
P. F. van der Meer & R. N. I. Pietersz
Sanquin Blood Bank North-West Region, Amsterdam, The Netherlands
Background and Aim
(BTHC-PVC) containers were compared for storage of leukoreduced platelet
concentrates (LR-PC): three ethylene oxide (EtO) sterilized (Gambro, Haemonetics,
Fresenius), and one steam-sterilized (Fresenius).
Methods
LR-PCs were made from 5 buffy coats and 300 ml Composol additive
solution, and leukoreduced by filtration. Four LR-PCs were pooled and subsequently
divided over the 4 BTHC-PVC bags to prevent donor-dependent differences, and
sampled for
in vitro
analysis on day 1, 2, 5, 7 and 9.
Results
The pH values on day 9 were (mean
7·12
±
0·04 (Haemonetics), and 7·07
±
0·09 (Fresenius, EtO-sterilized) (not significantly
different), vs. 6·91
±
0·12 (Fresenius, steam-sterilized;
bags). LR-PCs stored in the steam-sterilized bag exhibited significantly higher glucose
consumption and lactate production (
P <
0·001 vs. all EtO-sterilized bags).
Conclusion
All BTHC-PVC containers allow storage of LR-PCs for up to 9 days with
good
in vitro
parameters. However, the method of sterilization affects the storage
conditions of the LR-PCs in these bags.
Key words:
ethylene oxide, extension of platelet storage, platelet concentrates, steam
sterilization, sterilization.
Four butyryl trihexyl phthalate plasticized polyvinyl chloride
±
SD,
n
= 10): 7·12
±
0·03 (Gambro),
P
< 0·001 vs. all EtO-sterilized
Received: 28 July 2006,
revised 12 September 2006,
accepted 26 September 2006,
published online 31 October 2006
Introduction
In vitro
affected by a number of variables, such as the type of plastic
of the storage container, the thickness of the plastic film, and
the nominal volume of the container, giving the overall gas
permeability of the storage container. Further, the platelet
concentration and/or platelet number per container, the
storage period, and the type of platelet storage medium used
(plasma or additive solutions) influence the storage
characteristics of a PC.
Both the buffy coat (BC) method for preparation of PCs, and
single-donor apheresis PCs contain high numbers of platelets,
requiring storage containers that adequately expel carbon dioxide,
and allow entry of sufficient oxygen. This has led to the develop-
storage parameters of platelet concentrates (PC) are
ment of containers with a large ‘breathing’ surface (usually
expressed as the ‘volume’ of a storage container), and these
are made of special plastics with high gas permeability [1–3].
One such plastic is butyryl trihexyl citrate (BTHC) plasti-
cized polyvinyl chloride (PVC). We have shown earlier that
containers made from this plastic allow at least 12-day
storage of leukoreduced platelet concentrates (LR-PC) with
maintenance of
in vitro
quality [4]. Various manufacturers
use this plastic for production of platelet storage containers,
both for BC-derived and apheresis PCs. BC pooling systems
can be sterilized dry by gamma irradiation or wet by steam,
but this can result in small shape alterations of the plastics.
Therefore, because apheresis systems have moving parts,
ethylene oxide (EtO) sterilization of the whole set including
storage containers is preferred for those applications.
The aim of this study was to evaluate BTHC-PVC storage
containers from various manufacturers for storage of BC-
derived LR-PCs for up to 9 days after blood collection. We
evaluated the influence of the method of sterilization of the
storage containers on the platelet quality
storage. Following the results of the study, the gas permeability
in vitro
during
Correspondence
West Region, Plesmanlaan 125, 1066 CX Amsterdam, PO Box 9137, 1006 AC
Amsterdam, the Netherlands. Tel.: +31 20 5123281; Fax: +31 20 6178080;
E-mail: p.vandermeer@sanquin.nl
: Pieter F van der Meer, PhD, Sanquin Blood Bank North

Page 2

©
Journal compilation
2006 The Author(s)
©
2006 Blackwell Publishing Ltd.,
Vox Sanguinis
(2007)
92
, 32–36
Platelet storage container sterilization
33
of the plastic films of the containers in this study was
measured, and platelet activation was measured for two of
the containers in the main study.
Materials and methods
Whole blood collection and processing
Whole blood (500
systems (T2988, Fresenius HemoCare, Emmer-Compascuum,
The Netherlands) and cooled to 20–24
diol plates. Units were stored at 20–24
centrifuged (30 000
g
min). The whole blood was processed
into plasma, a red cell concentrate and a buffy coat (BC) using
the automated separator Compomat G4 (Fresenius). The BC
had a volume of 50 ml and a haematocrit of 40%.
±
50 ml) was collected in top and bottom
°
C using butane-1,4-
°
C overnight and
Preparation of platelet concentrates
PCs were prepared by pooling 5 AB0-identical buffy coats
and 300 ml Composol (300 ml, Fresenius) in a BC pooling
system (Terumo, Tokyo, Japan). This resulted in a 35% final
plasma concentration and 65% additive solution. The pools
were centrifuged, and the platelet-rich supernatant was
expressed through a leukoreduction filter to a container. Four
LR-PCs were pooled in a 2-l bag, and equally divided into
1·3-l BTHC-PVC containers from Gambro (art. no. 777-000-300,
EtO-sterilized), Haemonetics (art.no. 994CF-E, EtO-sterilized),
Fresenius (made for the purpose of this study, EtO-sterilized),
and Fresenius (art.no. T2104, steam-sterilized). The LR-PCs
were stored on a flatbed shaker (60 strokes/min) at 20–24
and sampled at day 1 (i.e. immediately after preparation of
the PCs), days 2, 5, 7 and 9 through a sample site coupler
using a needle and syringe (
n
= 10 paired experiments).
Similarly, to demonstrate platelet activation induced by
the storage containers, 2 LR-PCs in Composol were pooled
and divided into the EtO and steam-sterilized Fresenius
containers as described above, but with sampling on days 1,
6 and 8 (
n
= 4 paired experiments).
°
C,
In vitro measurements
Volumes were calculated from the net weight and specific
gravity (1·026 g/ml for plasma, and 1·006 for additive solution).
Cell counts and measurement of mean platelet volume (MPV)
were performed on a Sysmex K-1000 automated counter
(TOA, Tokyo, Japan). Leukocytes in LR-PCs were counted on
a flow cytometer (FacsCalibur, BD Biosciences, San Jose, CA,
USA) using the LeucoCount method according to the manu-
facturer’s instructions. Red cells were counted in a Bürker
haemocytometer in a 1:5 dilution in phosphate-buffered
saline. Blood gases, pH, bicarbonate, glucose and lactate were
measured on an ABL 705 blood gas analyser (Radiometer,
Copenhagen Denmark) at 37
a scale from 0 (poor swirl) to 3 (excellent swirl). The platelet
morphology was determined by microscopically judging 100
platelets on a scale from 0 (poor morphology) to 4 (excellent
morphology). CD62P expression was measured by labelling
formaldehyde-fixed platelets with fluorescein isothiocyanate
(FITC)-conjugated anti-CD62P (art.nr. 555523, BD Biosciences).
Fixed platelets incubated with FITC-conjugated anti-IgG1
(art.nr. 345815, BD Biosciences) served as negative control.
These stained samples were measured on the flow cytometer.
°
C. Swirl was scored visually on
Measurement of gas permeability and film thickness
Oxygen permeability was measured with an oxygen perme-
ability tester (Ox-Tran 100 A, Mocon, Minneapolis, MN,
USA) and carbon dioxide permeability with a programmable
permeability tester (Permatran C-IV, Mocon) according to the
manufacturer’s instructions, at 23
The films were conditioned by placing the samples 24 h in
the apparatus before measuring. Film thickness was measured
with a digital indicator (U12A, Sony Precision Technology
Europe, Stuttgart-Wangen, Germany).
°
C and 0% relative humidity.
Requirements
All tested PCs had to conform to the Dutch and European
guidelines for LR-PCs (whichever was more strict): a volume
of 150–400 ml, > 250
×
10 platelets, < 1·0
in 90% of the units and < 2
×
10
the pH (measured and reported at 37
6·8 and 7·4, and swirl should remain present.
9
×
10
6
leucocytes
9
red cells. Throughout storage,
°
C) should be between
Statistical analysis
Data are expressed as mean
using
INSTAT
(version 3·06, GraphPad, San Diego, CA, USA).
Storage data were compared using a repeated-measure
followed by a Tukey post test. Comparison of storage parameters
with the day 1 values was done using the Dunnett’s test. The
gas permeability of the containers was compared with an unpaired
, followed by a Tukey post-test. A
considered to indicate a statistically significant difference.
±
SD. Results were analysed
ANOVA
,
ANOVA
P
-value < 0·05 was
Results
The pooling and division of the LR-PCs was highly effective,
with no significant differences between the groups on day 1
(see Table 1). All LR-PCs conformed to our requirements.
During storage (see Fig. 1), LR-PCs in the Gambro and
Haemonetics containers both showed an initial rise in pH
between day 1 and 2, after which the pH remained constant
until day 9 of storage (day 2 through day 9 all
compared with the day 1 value). PCs in the EtO-sterilized
P
< 0·01 as

Page 3

34
P. F. van der Meer & R. N. I. Pietersz
©
2006 The Author(s)
(2007)
92
Journal compilation
©
2006 Blackwell Publishing Ltd.,
Vox Sanguinis
, 32–36
Fresenius container showed constant pH levels throughout
storage (day 2 through day 9 not significantly different from
day 1). In contrast, LR-PCs in the steam-sterilized Fresenius
bag showed a gradual decline in pH, reaching a statistically
lower pH value on day 9, as compared with the day 1 value
(
P <
0·01). At the end of the storage period, the LR-PCs in the
steam-sterilized containers had significantly lower pH values
than those stored in EtO-sterilized ones.
Oxygen levels were higher in LR-PCs stored in the Gambro
container compared with the others. There was a significant
difference between the carbon dioxide levels in all tested
containers. At the completion of storage, glucose levels were
Table 1 In vitro storage characteristics of leukoreduced platelet concentrates, stored in various BTHC-plasticized PVC containers (n = 10 paired experiments)
Manufacturer
Sterilization method
Fresenius
Steam
Fresenius
Ethylene oxide
Gambro
Ethylene oxide
Haemonetics
Ethylene oxide
Day 1
Volume, ml
Platelets, × 109
Leukocytes, × 106
Red cells, × 109
Day 9
pH
PO2, mmHg
PCO2, mmHg
Glucose, mM/L
Lactate, mM/L
Bicarbonate, mM/L
MPV, fL
Swirl
Morphology
355 ± 4
348 ± 19
0·00 ± 0·01
0·15 ± 0·10
351 ± 6
340 ± 20
0·02 ± 0·03
0·16 ± 0·12
352 ± 6
340 ± 18
0·02 ± 0·04
0·17 ± 0·10
352 ± 4
335 ± 23
0·02 ± 0·03
0·16 ± 0·10
6·91 ± 0·12a
58·8 ± 9·1
29·9 ± 1·1c
2·2 ± 0·8a
10·8 ± 1·2a
5·8 ± 1·1a
9·0 ± 0·4d
3 ± 0
211 ± 39e
7·07 ± 0·09
55·7 ± 16·4
27·9 ± 1·8c
3·2 ± 0·7
9·3 ± 1·0
7·8 ± 1·2
8·6 ± 0·2
3 ± 0
241 ± 22
7·12 ± 0·02
76·1 ± 10·4b
25·1 ± 1·3c
3·4 ± 0·3
9·1 ± 0·8
7·8 ± 0·4
8·8 ± 0·3
3 ± 0
228 ± 23
7·12 ± 0·04
61·2 ± 11·2
26·6 ± 1·4c
3·6 ± 0·4
8·7 ± 0·7
8·3 ± 0·5
8·7 ± 0·4
3 ± 0
236 ± 21
aP < 0·001 vs. all EtO-sterilized containers; bP < 0·001 vs. other containers; cP < 0·001 among all containers; dP < 0·05 vs. Fresenius and Haemonetics
EtO-sterilized; eP < 0·05 Fresenius steam vs. EtO-sterilized.
Fig. 1 Storage parameters of leukoreduced
platelet concentrates in Composol additive
solution, stored for up to 9 days in either
steam-sterilized Fresenius containers (?), or
ethylene dioxide sterilized containers from
Fresenius (?), Gambro (?) or Haemonetics (?)
(mean SD, n = 10 paired experiments).

Page 4

© 2006 The Author(s)
Journal compilation © 2006 Blackwell Publishing Ltd., Vox Sanguinis (2007) 92, 32–36
Platelet storage container sterilization
35
lower and lactate levels higher for LR-PCs in the steam-
sterilized Fresenius container. The resultant glucose con-
sumption rate was 0·18 ± 0·03 mmol/day/1011 platelets in
the steam-sterilized container, which was significantly higher
than in the LR-PCs in the Gambro, Haemonetics, and
EtO-sterilized Fresenius bags (0·14 ± 0·01, 0·14 ± 0·01 and
0·15 ± 0·03 mmol/day/1011 platelets, respectively, P < 0·001).
Lactate production was 0·30 ± 0·04 mmol/day/1011 platelets
for LR-PCs in the steam-sterilized Fresenius bag, signifi-
cantly higher than in the Gambro, Haemonetics, and
EtO-sterilized Fresenius bags (0·24 ± 0·03, 0·23 ± 0·03 and
0·25 ± 0·03 mmol/day/1011 platelets, respectively, P < 0·001).
Probably as a consequence of formation of acid meta-
bolites, bicarbonate levels were lower in the LR-PCs stored in
steam-sterilized containers. MPV and morphology scores
showed minor differences as compared with values of other
containers. The swirling score remained excellent in all units
on all tested days.
Following these results, gas permeability, container surface
and film thickness were measured, as summarized in Table 2.
Both oxygen and carbon dioxide permeability were signifi-
cantly lower for the EtO-sterilized Fresenius container as
compared with the other 3 containers. The Fresenius bags had
a larger surface and a thicker film than the Gambro and
Haemonetics containers. When correcting for the different
bag surfaces (assuming both sides of the bag contribute
equally to the total gas exchange capacity), the overall gas
exchange capacity of the steam-sterilized container was
significantly higher as compared with the EtO-sterilized
containers (Table 2).
Because these results could not explain our initial obser-
vations of the effect of sterilization, an additional storage
experiment was conducted, comparing only the steam and
EtO-sterilized Fresenius containers, to determine if platelet
activation caused the elevated glucose consumption and
lactate production. Indeed, a significantly higher CD62P
expression was found on day 8 for platelets stored in the
steam-sterilized container vs. the EtO-sterilized one (24·0 ± 3·2 vs.
19·5 ± 1·6%, respectively, P < 0·05). The other storage
parameters were in line with the data reported for the main
storage study (not shown).
Discussion
This study shows that all BTHC-PVC containers tested in this
study allowed up to 9 days storage of LR-PCs in Composol-
PS additive solution with maintenance of good in vitro
quality parameters, independent of the sterilization method.
However, in comparison to the EtO-sterilized containers,
LR-PCs stored in the steam-sterilized container showed
significantly lower pH values, higher glucose consumption
and lactate production rates, and lower bicarbonate concen-
trations following 9 days of storage, indicative of suboptimal
storage conditions.
We first hypothesized that the method of sterilization had
affected the characteristics of the plastic film. This phenomenon
is scarcely described in the literature. One study described
the effect of sterilization on containers for storage of
granulocyte concentrates [5]. That study showed that
tri-(2-ethylhexyl)trimellitate (TOTM) plasticized PVC bags
sterilized with steam gave significantly lower pH values as
compared with EtO-sterilized bags. It was established that steam
sterilization reduced the CO2 permeability by 10%. This, in com-
bination with a 5–10% shrinkage of the film following steam
sterilization was used as explanation for the observed lower pH
values. In our study, however, using a different plastic, the
surface area or film thickness was not affected by the method
of sterilization. Subsequent measurement of oxygen and
carbon dioxide permeability showed lower gas permeability
for the EtO-sterilized Fresenius container as compared with
the steam sterilized one. So, different gas permeabilities induced
by the sterilization could not explain our observations.
In a more recent study [6], comparing EtO- and steam-
sterilized BTHC-plasticized PVC containers for storage of platelet
concentrates, no differences in platelet storage characteristics
were observed. In that study, apheresis platelet concentrates
collected in acid citrate dextrose solution A (ACDA) antico-
agulant were studied, as compared with buffy-coat-derived
Table 2 Gas permeability and thickness of BTHC-PVC plastics used for platelet storage containers (shown as average of a duplicate measurement)
Manufacturer
Sterilization method
Fresenius
Steam
Fresenius
Ethylene oxide
Gambro
Ethylene oxide
Haemonetics
Ethylene oxide
Film thickness, µm
Oxygen permeability, cm3/m2/day/bar
CO2 permeability, cm3/m2/day/bar
Surface, cm2
Oxygen exchange capacity, cm3/day/bar
CO2 exchange capacity, cm3/day/bar
416
2234
15 253
372d
166e
1135
405
2035a
14 279c
373
152
1065
382
2283
16 565
331
151
1097
386
2126b
15 794
341
145
1077
aP < 0·05 vs. Fresenius steam and P < 0·01 vs. Gambro; bP < 0·01 vs. Gambro; cP < 0·01 vs. Gambro and < 0·05 vs. Haemonetics; dP < 0·05 vs. Gambro and
Haemonetics; eP < 0·01 vs. all EtO-sterilized containers.

Page 5

36
P. F. van der Meer & R. N. I. Pietersz
© 2006 The Author(s)
Journal compilation © 2006 Blackwell Publishing Ltd., Vox Sanguinis (2007) 92, 32–36
platelets in CPD in our study. Any of the following three
variables – or a combination of these – could explain the
differences between their and our study: (i) apheresis vs.
buffy coat-derived platelets [7]; (ii) ACDA vs. CPD; the latter
contains phosphate, which is known to stimulate lactate
formation [8]; and (iii) plasma vs. Composol; Composol has
no buffering capacity. Only one in vitro parameter showed
an effect [6], which was a higher expression of the activation
marker CD62P on platelets that were stored in the steam-
sterilized containers. The latter result, in combination with
a previous observation that some platelet containers showed
poor storage conditions that was attributed to a rough grid
pattern of the plastic film [3], suggested that platelets in the
steam-sterilized containers were activated by characteristics
of the plastic film, inducing platelet activation. Therefore, a
second storage study comparing only the steam- and EtO-
sterilized Fresenius containers was performed. We found a
significantly higher CD62P expression in the steam-sterilized
containers as compared with the EtO-sterilized ones,
supporting our hypothesis that the method of sterilization
affects the properties of the plastic film used for the platelet
storage container. This causes activation of platelets stored
in steam-sterilized containers, shown as a higher CD62P
expression and elevated glucose and lactate metabolism.
Finally, throughout the 9 days of storage in the current
experiments, a minimal change of in vitro parameters was
observed for the platelet concentrates in the EtO-sterilized
containers. Although it is tempting to claim that extension
of the storage period is feasible, clinical studies should be
performed to provide sufficient evidence to allow extended
storage. One such clinical study was performed by our centre
for platelet concentrates in plasma, and 7-day storage is
currently allowed in our country [9].
We conclude that BTHC-PVC bags allow at least 9-day
storage of LR-PCs with maintenance of pH > 6·8. The method
of sterilization influences the in vitro storage conditions of
LR-PCs, possibly by altering the surface of the storage
container, inducing platelet activation and elevated platelet
metabolism. Therefore, we caution that the method of
sterilization can influence the storage characteristics of
platelet concentrates, and should be taken into account when
choosing and/or validating a new platelet storage container
in blood bank routine.
Acknowledgements
We thank Lara Liefting for excellent technical assistance. We
thank Renolit Nederland BV (Enkhuizen, The Netherlands)
for performing gas permeability measurements at no cost.
References
1 Gulliksson H, Shanwell A, Wikman A, Reppucci AJ, Sallander S,
Uden AM: Storage of platelets in a new plastic container.
Polyvinyl chloride plasticized with butyryl-n-trihexyl citrate.
Vox Sang 1991; 61:165–170
2 De Wildt-Eggen J, Schrijver JG, Bouter-Valk HJ, Fijnheer R, Bins M,
van Prooijen HC: Improvement of platelet storage conditions by
using new polyolefin containers. Transfusion 1997; 37:476–481
3 Van der Meer P, Pietersz R, Reesink H: Leucoreduced platelet
concentrates in additive solution: an evaluation of filters and
storage containers. Vox Sang 2001; 81:102–107
4 Van der Meer PF, Pietersz RN, Reesink HW: Storage of platelets in
additive solution for up to 12 days with maintenance of good
in-vitro quality. Transfusion 2004; 44:1204–1211
5 Miyamoto M, Sasakawa S: Effects of autoclave sterilization on
the physical properties of storage bags and granulocyte function.
Vox Sang 1988; 54:74–77
6 Ramirez M, Dumont LJ: Steam sterillization of platelet storage
bags – effects on in vitro characteristics. Transfusion 2003;
43(Suppl.1):SP75 (Abstract).
7 Vassallo RR, Murphy S: A critical comparison of platelet prepara-
tion methods. Curr Opin Hematol 2006; 13:323–330
8 Gulliksson H, Larsson S, Kumlien G, Shanwell A: Storage of
platelets in additive solutions: effects of phosphate. Vox Sang
2000; 78:176–184
9 Dijkstra-Tiekstra MJ, Pietersz RN, Hendriks EC, Reesink HW,
Huijgens PC: In vivo PLT increments after transfusions of WBC-
reduced PLT concentrates stored for up to 7 days. Transfusion
2004; 44:330–336