Vox Sanguinis (2014) 106, 209–218

© 2013 International Society of Blood Transfusion

REVIEW DOI: 10.1111/vox.12089

Residual aggregates in platelet products: what do we

know?
J. Ringwald,1 M. Antoon,2 R. Eckstein1 & M. Cardoso2
1
Department of Transfusion Medicine and Hemostaseology, University Hospital of Erlangen, Erlangen, Germany
2
Terumo BCT Biotechnologies, Zaventem, Belgium

Received: 4 February 2013,
revised 11 July 2013,
accepted 18 July 2013,
published online 9 October 2013
Background and Objectives Platelet (PLT) aggregates can occur during or after
PLT component processing. However, very few reports investigating the pheno-
menon and its clinical significance have been published. In this review, currently
available information about aggregates in PLT products is summarized and possi-
ble causal factors as well as preventive strategies are discussed.
Materials and Methods A review of the MEDLINE
database for relevant publi-
cations from 1960 to May 2013 was conducted.
Results It is well known that PLT aggregates may occur during or after the PLT
product preparation process. These aggregates normally dissipate with rest and
agitation. However, in some rare cases, the aggregates do not dissipate within
24 h and can persist up to the end of storage. Exposure to low temperature, low
pH, short resting period after collection, different collection systems, presence of
bubbles or foam inside the PLT bag, PLT-container interactions, proper product
mixing and donor-dependent variables may have an impact on the formation of
PLT aggregates. Although publications are rare, the presence of small numbers of
PLT aggregates appears to have only limited impact on PLT in vitro quality.
Furthermore, data on the clinical impact of PLT aggregates are lacking.
Conclusion Despite the fact that PLT aggregates occur in PLT products, published
data on this topic remain scant. Considering the concern of clinicians about this
phenomenon, more studies are needed which should focus on the possible clini-
cal impact of such aggregates and precautions to avoid PLT aggregate formation
in PLT products.
Key words: apheresis platelets, platelet aggregates, platelet concentrates, platelets
components, whole blood-derived platelets.

example, at the blood centre when staff notice irreversible

Introduction
Platelet (PLT) aggregate formation in whole blood (WB)-
and apheresis-derived PLT products has been known to
occur, and their prevention was a point of high interest
when component technologies were first developed [1].
Products with macroaggregates or clumps may be dis-
carded due to concerns linked to efficacy and safety of
the product. Discard may occur at different occasions, for
Correspondence: Juergen Ringwald, Department of Transfusion
Medicine and Haemostaseology, University Hospital of Erlangen,
Krankenhausstr. 12, D-91054 Erlangen, Germany
E-mail: juergen.ringwald@uk-erlangen.de
clumps, in the hospital blood bank when clumps are
missed by the blood centre or form on blood bank
shelves, or on the hospital ward when clumps are missed
by the blood bank staff or form during transport to the
ward [2, 3]. Currently there is no consensus on a general
discard strategy and procedures therefore vary from cen-
tre to centre. Some centres may discard all products with
any residual aggregates at 24 h, and other centres might
further manufacture clumped units by (leuco) filtration
and discard units still containing aggregates after filtra-
tion process or others release products only with less than
a prespecified number of microaggregates but no products
containing macroaggregates (personal communication

209
210 J. Ringwald et al.

with staff of different blood centres). Already in the early
days of manufacturing of PLT components, further acidi-
fication of PLT products by adding citric acid was intro-
duced to prevent this phenomenon [4]. However, the
resulting lower pH was a significant barrier to storage of
the PLT products for longer than 24 h as further decrease
in the pH below 6
0 is associated with decreased PLT
quality [5,6]. PLT aggregates are generally described as
groups or clusters of PLTs.
The kind and appearance of clumping as well as the
causes responsible for this phenomenon might differ
between WB- and apheresis-derived PLT products mainly
due to different shear forces and PLT activation induced
by the various techniques. In the early days of the prepa-
ration of buffy-coat-derived PLT concentrates, clumps
had been observed in the buffy coats after initial centrifu-
gation, and longer storage of those buffy coats seemed to
be associated with a higher PLT activation level [7]. In
contrast, PLT aggregation in apheresis-derived PLT prod-
ucts – sometimes referred to as ‘postpump clumping’ or
‘friendly clumping’ – may also be related to the exposure
of PLTs to centrifugal forces and shear stress (pumps)
during collection which may lead to increased activation
and possibly aggregation [8, 9]. For both kinds of PLT
products, however, temperature changes during the man-
ufacturing process may be of relevance with regard to the
formation of clumps. For instance, the temperature during
collection and storage of WB or of PLT pellets prior to
PLT resuspension may influence PLT aggregability in
WB-derived PLT products [10,11]. PLT aggregates may
dissipate during rest periods and agitation [8]. However,
in some rare cases, the aggregates do not dissipate within
24 h postcollection and can persist longer even through-
out the storage period [9]. Although aggregates occur reg-
ularly in PLT products, their relevance in terms of their
impact on product quality, clinical efficacy and safety
has never been systematically studied and publications
referring to this aspect of PLT quality are scarce and often
contain very limited information. Therefore, in an effort
to address some of the questions raised by blood bankers
relative to this topic and to determine steps for further
follow-up investigations, this review collates the informa-
tion that is currently available.
the phenomenon of aggregates as a quality aspect of PLT
products as primary or secondary subject matter were
withheld for inclusion in the review. Furthermore, the
individual reference lists of all found materials that were
selected in this manner were scanned for additional
papers.
Results
Defining PLT aggregates
While systematic studies on the subject of residual aggre-
gates in PLT products do not exist, as mentioned above,
Devine et al. [12] looked into the subject in response to
reports of an increase in PLT particulate matter seen after
the nationwide introduction of universal leucoreduction
in Canada. In general, macroscopically visible PLT aggre-
gates are defined as clumps of white amorphous matter
ranging in size from less than one to several millimetres
(mm) in diameter. A further distinction can be made on
the basis of the persistence of aggregates between
‘friendly’ reversible and residual aggregates (Fig. 1). The
first group is characterized by their larger size (>1 mm).
They are relatively common and tend to dissipate during
rest periods and agitation within 24 h postcollection. The
second group of residual aggregates are characterized by
their smaller size (<1 mm), often no more than the size of
a pinhead, and are often referred to as microaggregates.
They are less frequent, but when they occur, they may
(a)
(b)

Methods
A review to identify relevant literature by means of the
MEDLINE
database was performed. A period ranging
from 1960 until May 2013 was researched. The following
key words and text words were used as search terms in
combination with ‘PLT aggregates’ or ‘PLT clumping’:
‘PLT concentrate(s)’, ‘PLT apheresis’, ‘PLT transfusion’ and Fig. 1 Example for reversible (friendly, postpump) macroaggregates (a)
‘whole blood-derived PLT(s)’. Publications that addressed and for residual microaggregate (b).

© 2013 International Society of Blood Transfusion

Vox Sanguinis (2014) 106, 209–218

persist up to the end of storage [9]. In some cases,
depending on the standard operating procedure of the
blood bank, these microaggregates may prevent the prod-
ucts from being released for transfusion and thus lead to
product loss. Aside from the visual definition of a PLT
aggregate, the question of the exact nature and composi-
tion of aggregates has not been extensively investigated.
According to early reports, proteins such as fibronectin
were found to be the main constituent of the particulate
matter detected in PLT products [13–15]. More recently,
other groups reported that the particulate matter found in
some of their PLT products consisted of PLTs and PLT
aggregates [11,12,16]. Whether this discrepancy could be
explained by the fact that the more recently published
data identifying PLTs or PLTs aggregates as the main
constituents of the particulate matter investigated filtered
PLT concentrates has been discussed by Moroff et al. [16]
but remains not totally clear. However, Devine et al. [12]
isolated and analysed clumps in their natural history most
extensively by means of light microscopy, electron
microscopy, protein electrophoresis, and protein and
phospholipid analysis. They found that all aggregated
material was PLT derived without any traces of fibronec-
tin and no evidence of leucoreduction filter fibres.
Causes of aggregate formation in PLT products
In early work studying the impact of citrate solutions
with different pH levels on the clumping tendency of PLT
products, Aster et al. [17] found that clumping was pre-
vented at pH 6
5 and they found that a citrate solution
containing sufficient acid lowers the clumping. At the
same time, Flatow et al. [4] found similar results for PLT
concentrates prepared in acidified plasma (pH 6
5–6
7)
that were superior in terms of PLT recovery and post-
transfusion increment compared with PLT components
prepared by standard methods. These initial observations
about the link between pH and PLT aggregation founded
the current anticoagulation strategies for WB- as well as
apheresis-derived PLT products. One key factor in opti-
mizing the use of anticoagulant (AC) is the aspect of
proper mixing or resuspension during the manufacturing
process. With regard to WB-derived PLT products, inade-
quate mixing of WB with the AC during collection might
already result in more activated PLTs. For PRP-derived
PLT products the smooth and complete resuspension of
the PLT pellets after centrifugation has been of special
interest since the early days of this method [1]. The
importance of proper resuspension in preventing
unwanted aggregation was discussed in a paper by
Snyder et al. [18]. They described how a more intensive
lateral movement (1
5 vs. 1
0 inch) improved the resus-
pension of PLT concentrates. However, the resuspension
© 2013 International Society of Blood Transfusion
Vox Sanguinis (2014) 106, 209–218
Aggregates in platelet products 211
of highly concentrated PLTs is also of importance in some
distinct apheresis techniques as will be discussed further
below.
Besides the pH and mixing of PLT concentrates,
another established factor with important impact on PLT
aggregation was found to be the storage temperature. The
increased PLT stickiness induced by cooling of PLT sus-
pension has been known for a long time [19]. Already in
1967, Pert et al. [20] reported more clumping in PLT con-
centrates prepared from chilled blood than in those pre-
pared from blood kept at room temperature. In 1971,
Kattlove et al. [21] tested aggregation in PLT concentrates
at different temperatures (0–1, 6, 23 and 37°C) and found
that PLTs stored at or exposed to low temperatures (0–1
and 6°C) showed significantly higher aggregation ten-
dency. They postulated that since protein conformation is
affected by changes in temperature, cold may induce PLT
aggregation by causing conformational changes in PLT
membrane protein. In 1985, Welch et al. [10] reported the
successful reduction in the number of clumped units
when the temperature of the laboratory, centrifuge and
storage environmental chamber was increased to 24°C.
More recent research has shown that temperature during
processing, even within the accepted 20–24°C range, also
significantly affects aggregate formation and that raising
the temperature from 20 to 22°C to 24°C significantly
reduces the formation of macroaggregates [11].
Another factor often mentioned in connection to PLT
aggregation is resting. PLT-containing storage bags are
best kept (label-side down) on a laboratory work bench at
room temperature away from draft or cold air flows
before being placed on the agitator [22, 23]. Skripchenko
et al. [8] investigated the effect of different rest times
without agitation on a set of in vitro quality parameters
and the formation of PLT aggregates. However, the study
was unable to establish any links between resting time
and presence of PLT aggregates, as macroscopically visi-
ble PLT aggregates observed in five of 15 products had
all dissolved by day 1 already. The investigators did find
a positive impact of a longer resting time on other
in vitro parameters such as the extent of shape change
(ESC), percentage of discoid PLTs and expression of GP1b-a.
The size of the PLTs has also been found to be corre-
lated to the aggregation potential of PLTs. In a study
published in 1989, Wong et al. [24] proposed that large
PLTs are intrinsically more sensitive and more rapidly
recruited into aggregates. PLT size can vary considerably
between individual donors and can depend on multiple
factors such as underlying disease and metabolic differ-
ences [25].
During recent years, collecting plasma-reduced PLT
concentrates by apheresis and the use of PLT additive
solution (PAS) as a resuspension medium has become
212 J. Ringwald et al.
more popular. To our knowledge, there is currently no
clear evidence in the literature as to whether using PAS
as a storage medium in PLT production is generally asso-
ciated with a higher incidence of PLT aggregates in PLT
concentrates. A possible reason for such an observation
could be the lower content of albumin or other plasma
proteins which are known to have PLT aggregation inhibi-
tory effects [26–28]. Compared with PLT concentrates col-
lected and stored in 100% plasma, PLT concentrates
containing 60–70% of a rather simple PAS such as PAS-II
or PAS-III showed inferior in vitro PLT quality with
increased CD62P-expression indicating a higher PLT acti-
vation level [29]. To further increase the in vitro quality
of PAS-stored PLTs concentrates, the inclusion of magne-
sium and potassium chloride was suggested as several
authors had described an inhibitory effect of magnesium
and/or potassium on PLT activation and aggregation
[30–32]. De Wildt-Eggen et al. [33] were the first system-
atically studying the effect of the addition of magnesium
and potassium on PLT’s CD62P-expression. Using PAS-II
as basic PAS, they found significantly lower levels of
CD62P-expression in those PLT components containing
additional magnesium and potassium. Other groups
confirmed their findings for PAS-III supplemented with
magnesium and potassium as being also associated with a
significantly improved in vitro PLT quality in general
[34–36]. These results have led to the inclusion of these
ions in the recent generation of PAS [37]. However, the
role of these ions in lowering the PLT activation level has
been questioned by unpublished personal reports from
blood banks observing an increase in PLT aggregation
upon switching from PAS solutions without magnesium
and potassium such as T-Sol
or Intersol
to solutions
containing these two ions. However, it is difficult to dis-
sect whether this observation is an unexpected effect of
the new generation of PAS or rather a phenomenon that
can be attributed to the enhanced surveillance and scru-
tiny that novel blood components or processes are gener-
ally exposed to before routine implementation.
Some studies have reported differences in the frequency
of residual aggregation depending on the method of PLT
preparation. However, process-induced PLT activation as
probably the most important cause for the formation of
aggregates may occur with all production methods [38].
Compared with PRP-derived PLT concentrates, Vasconcelos
et al. [39] observed less evident PLT aggregation in buffy-
coat- and apheresis-derived PLT concentrates. This seems
to be in agreement with observations of other groups who
found significantly higher PLT activation levels for PRP-
derived compared with buffy-coat-derived PLT products
[40,41]. For apheresis-derived PLTs collected with the
Haemonetics PCS-Plus (Haemonetics Ltd.), Metcalfe et al.
[40] reported PLT activation levels in-between those of
PRP- or buffy-coat-derived PLT products. The presence of
aggregates as well as the PLT activation levels seems to
vary depending on the apheresis technique used. Compar-
ing four apheresis devices (CS-3000 Plus
, Fenwal; MCS-
3p
, Haemonetics; Amicus Cresendo
, Fenwal; Spectra
,
Terumo BCT), Nakajo et al. [42] found a significantly
higher frequency of initial PLT clumps in PLT components
collected with the CS-3000 Plus (30
5%) and the Amicus
(22
2%) compared with MCS-3p (6
4%) and Spectra (9
8%).
However, within 72 h of storage, nearly all clumps disap-
peared except for a low residual rate of 0
8% in PLT prod-
ucts collected with CS-3000 Plus or Amicus [42]. The
higher PLT activation level of PLTs in components col-
lected with Amicus compared with Spectra and MCS+ was
confirmed by Hagberg et al. [43] just 1 year later. More
recently, Flesch et al. [9] compared the two apheresis plat-
forms Amicus and Trima Accel
(Terumo BCT) and found
reversible micro- and macroscopically visible aggregates in
11% (10/91) of the Amicus and 1% (1/91) of the Trima, and
1/91 (1
1%) of irreversible aggregates in Amicus vs. 0/91
(0%) in Trima-derived PLT concentrates. Similar results
were reported by Skripchenko et al. [8]. These investigators
describe that in a total of 15 apheresis PLT units collected
on Amicus, five contained aggregates immediately follow-
ing collection. The aggregates were reversible and had all
dispersed by day 1 of storage. In agreement with Flesch
et al. [9], the authors point out that during apheresis col-
lection, PLTs are subjected to centrifugal forces and shear
stress which may lead to increased activation and possibly
aggregation. Therefore, the configuration of the optimal
ratio between AC and WB entering the system as well as
the design of new separation chambers to prevent clotting
by PLT aggregates are two major aims in the development
of apheresis technologies [44,45]. In the Amicus, the har-
vested PLTs are stored in a small bag within the centrifuge
during the entire apheresis procedure and are in intimate
contact with each other and the plastic surface of the col-
lection bag [43]. The manual resuspension required upon
completion of the procedure can, in some cases, remain
incomplete and result in persistent nondispersing aggre-
gates causing a small fraction of PLT units to be discarded.
In contrast, during plateletpheresis procedures using the
Spectra, MCS+ or Trima Accel, the collected PLTs are trans-
ferred continuously or intermittently to larger PLT storage
containers outside of the centrifuge. Interestingly, for the
precursor of the Trima Accel, the Trima with software, ver-
sion 4 (Trima V4), a very high PLT activation level was
described when collecting hyperconcentrated PLTs for stor-
age in PAS [36]. It was discussed that the high CD62P
expression might have been due to the different dual-stage
separation channel in the Trima v4 which contained a dam
where PLTs had been highly accumulated and brought in
contact with plastic surface [46]. In Trima Accel, the
© 2013 International Society of Blood Transfusion
Vox Sanguinis (2014) 106, 209–218

dual-stage separation channel was replaced by a single-
stage channel without such a dam, and CD62P levels were
significantly lower when collecting hyperconcentrated
PLTs [46].
Another factor potentially contributing to the presence
of macroaggregates in PLTs is the presence of air bubbles
and foam in the PLT bag. In a study by Sandgren et al.
[47] samples were aliquoted and prepared for storage with
(test) or without (reference) air bubbles/foam. The meta-
bolic, cellular and activation parameters as well as the
presence of visible aggregates of all units were analysed
during the 7-day storage period. Visible aggregates
occurred in three of the test units, but in none of the ref-
erence units. The higher rate of visible aggregates in the
test units was accompanied by an increased disintegration
of PLTs with overall decreased in vitro quality in these
units. Finally, Snyder et al. highlighted yet another aspect
linked to the PLT preparation process that may have an
important repercussion on the occurrence of PLT aggre-
gates in the stored product. They found that bags made
of materials with higher bag wall elasticity tended to
facilitate resuspension of aggregates present in the prod-
uct more easily [18].
Whether prestorage WBC reduction in PLT products
with the exposure of PLTs to negatively charged filter
material is associated with a significantly increased PLT
activation or PLT aggregate formation seems to be
unclear. Devine et al. [12] found significantly increased
PLT activation in their PLT concentrates derived from
PLT-rich plasma (PRP). They concluded that this could
explain their observation of an increased percentage of
PLT concentrates containing particulate material that
turned out to be PLT aggregates (see above) since prestor-
age WBC reduction in Canada was implemented. How-
ever, other groups did not find any alterations of PLT’s
activation status or formation of PLT aggregates when
pooled PLT concentrates or buffy-coat-derived PLT prod-
ucts were filtered [40,48,49]. Nonetheless, differences in
the production methods other than filtration could be also
responsible for the different observations. For instance,
Sweeney et al. [49] stated that they did not include PLT
units containing already visible clumps in their PLT pools
which may have contributed to the observed lack of mac-
roscopic aggregates in their pooled PLT concentrates. Con-
versely, WBC reduction filters have been shown to remove
activated PLTs expressing CD62P [50]. Overall, the well-
known benefits of prestorage WBC reduction might
outweigh a possible but not definitely proven increased
risk of PLT aggregate formation in filtered PLT products.
Finally donor-related parameters are often mentioned
as plausible cause for PLT aggregation. Jimenez et al.
[51] found differences in the potential to form
microaggregates between donors, which the investigators
© 2013 International Society of Blood Transfusion
Vox Sanguinis (2014) 106, 209–218
Aggregates in platelet products 213
successfully correlated to the efficacy of the transfusion.
According to the authors, donor quality reflects an intrin-
sic, donor-specific sensitivity of PLTs to functional degra-
dation caused by preparation and storage conditions
including reduction in extracellular calcium ion concen-
tration by citrate, changing temperature and pH, atypical
shear and isolation from regular contact with endothe-
lium. In our recently performed retrospective analysis
with regard to the occurrence of PLT aggregates in apher-
esis-derived PLT concentrates stored in 100% plasma, we
observed that PLT concentrates containing aggregates
were donated more often by females or donors with a
lower haematocrit in general [2]. A possible hypothesis
for this observation could be the relative higher plasma
content and therefore an increased need of AC to prevent
aggregate formation during apheresis process. However,
prospective studies are needed to clarify this hypothesis.
Although pooling of WB-derived PLT concentrates is
usually done using buffy coats or PLT concentrates of
identical ABO type, Sweeney et al. [52] mixed PLTs of A
and O type and compared in vitro qualities parameters
and the presence of microscopic clumps. They found
increased microscopic clumping and activation in the
mixed units, but PLT counts and other in vitro quality
parameters such as hypotonic shock response (HSR), ESC,
pH or morphology score were similar.
Finally, so called biofilms formed by bacteria such as
Serratia marcensens have been found to contain PLT
aggregates due to bacteria-induced PLT activation and
aggregation [53, 54].
Why some aggregates do not dissolve over time
remains unclear. A hypothesis may be different activation
levels of PLTs in reversible and persistent aggregates. The
PLTs in reversible aggregates may not be fully activated
but rather clump together without expressing GPIIb/IIIa
and binding via fibrinogen. PLTs in persisting aggregates
may be fully activated and exhibit shape change.
In summary, PLT activation may be one reason for the
formation of clumps or PLT aggregates but appears to be
not the only one. Therefore, PLT clumping or the occur-
rence of PLT aggregates seems to be multifactorial.
Table 1 lists the main factors that can potentially con-
tribute to aggregate formation in PLT products.
Aggregates and in vitro test parameters
Devine et al. [12] also looked at the in vitro characteristics
of PLT products with aggregates as compared to products
without. They found that the level of CD62P expression
(P-Selectin) correlated poorly with the degree of aggregate
formation. A similar finding was made when testing the
HSR: there was no difference between those units that
contained PLT aggregates and those that did not.
214 J. Ringwald et al.

Table 1 Different parameters with potential impact on formation of platelet (PLT) aggregates in PLT concentrates

Parameter Impact

Anticoagulant Increasing the anticoagulant to product ratio decreases the aggregation tendency [4, 17]
Resuspension of PLT pellets Incomplete resuspension of PLT pellets increases formation of PLT aggregates [1, 18, 47]
during preparation
(Storage) Temperature Exposure to low temperatures (<20°C) increases aggregation tendency [10, 11, 21]
Resting Correct and sufficient resting before agitation may lead to less aggregation tendency [8, 16]
Platelet size Larger platelets have been found to be more sensitive to aggregation and to be recruited
more rapidly in aggregates [24]
Platelet additive solution (PAS) Higher PLT activation levels were found in PLT products containing rather simple PAS [29].
Clear evidence for a generally increased aggregate formation in PAS-containing PLT
products is lacking.
Preparation technology Some component preparation technologies were found to be related to higher aggregation
tendency (e.g. PRP-method) [8, 9, 36, 38–42, 46]
Donor-related factors Interindividual differences in aggregate forming potential between donors [2, 51]
Air bubbles in PLT concentrates The presence of air bubbles due to insufficient evacuation of air trapped in the PLT product
facilitates PLT aggregate formation due to increased PLT desintegration [47]
Bacterial contamination In PLT components with positive bacterial growth bacteria-induced PLT activation and
aggregation may be found [53, 54]

Furthermore, products containing aggregates displayed
levels of C3a similar to those without aggregates, and there
was no correlation between the detection of thrombin acti-
vation markers and the presence of PLT aggregates. The
investigators concluded that the characteristics of the
remaining unaggregated PLTs, as well as the plasma pro-
teins, were not altered by the presence of aggregates.
Recently, a study was published that looked at the effect of
resting on aggregate formation in PLT concentrates [8]. The
authors compared the in vitro quality of PLTs derived from
PLT products with and without aggregates. In accordance
with Devine et al.[12], they found no significant differ-
ences for the ESC, morphology scores and GP1b (CD42b)
expression between units with or without aggregates after
collection. Even more recently, Feys et al. [55] reported
similar in vitro PLT characteristics in terms of aggregation
ability and activation levels for apheresis-derived PLT con-
centrates with or without visible particles. However,
approximately one decade before, Sowemimo-Coker et al.
[11] published an abstract indicating a somewhat decreased
in vitro quality of PLTs in units with aggregates with regard
to a slightly higher PLT activation status, lower HSR values
and lower PLT content. When comparing clump formation
in apheresis-derived PLT concentrates collected with differ-
ent devices, Nakajo et al. [42] detected an inverse relation-
ship between the time period clumps persisted and PLT’s in
vitro quality. Their data suggested that PLT concentrates that
contain clumps for at least 72 h could be of poor quality.
Finally, PLT aggregation may interfere when counting
residual leucocytes using flow cytometry. Figures 2a–c
show dot plots illustrating the potential consequence of
aggregates in the samples on the gating strategy for
leucocyte detection. The PLT aggregates form the black
tail on the diagonal of the plot. PLTs are autofluorescent
and fluoresce almost equally in the 9 and y direction
(without any stain). In the plot, the tip of the PLT tail is
included in the gated area R2 containing leucocytes, arti-
ficially increasing the residual leucocyte count.
Aggregates and clinical relevance
Very little evidence exists about the clinical relevance of
aggregates in PLT products. Some general considerations
can be made, however. In most countries, local regu-
lations require transfusion sets to contain a standard filter
with a pore diameter of 170–230 lm that will remove the
larger visible aggregates from the products upon adminis-
tration as shown very recently by Feys et al. [55]. How-
ever, smaller aggregates may be able to pass the filter.
To our knowledge, the negative effect of PLT aggregates
on product yield has not been investigated so far. How-
ever, Jimenez et al. [51] showed that PLT recovery of PLT
products derived from donors with an increased likelihood
to form citrate-induced microaggregates in an in vitro
functional test performed prior to apheresis was only one-
third of that in donors with no likelihood to form micro-
aggregates. The clinical relevance of this observation with
regard to the presence of visible macroaggregates in PLT
products and clinical efficiency remains unsure so far.
However, a simple calculation might illustrate that the
impact of the number of PLTs captured in the aggregates
appears to be negligible. Assuming an extreme scenario of
100 visible spherical PLT aggregates, each measuring
1 mm diameter (100 aggregates represent 50 9 109 fL;

© 2013 International Society of Blood Transfusion

Vox Sanguinis (2014) 106, 209–218

(a)
(c)
Fig. 2 Impact of platelet (PLT) aggregates on
counting of residual leucocytes by means of
flow cytometry. (a) dot plot of PLT sample with
aggregates – incorrect gating; (b) dot plot of
PLT sample with aggregates – correct gating;
(c) dot plot of PLT sample without aggregates.
100 9 Vsphere = 4/3p(0
5)3 = 50 mm3 = 50 9 109 fL) and
with a low mean PLT volume of 7 fL, this may represent
7
1 9 109 PLTs corresponding to 2
4% of all PLTs in a
PLT product containing 3 9 1011 PLTs. The impact of the
number of PLTs captured in the aggregates appears there-
fore to be negligible which was also found by others [16].
In addition, no cases have been reported in the litera-
ture where adverse transfusion events could be correlated
to the presence of aggregates in leucoreduced PLT prod-
ucts. However, this absence of reports on this subject may
not necessarily reflect the absence of problems, as this
may be due to the mere lack of awareness of a possible
relationship. Moreover, with the current haemovigilance
systems, it would be difficult and unlikely to relate any
clinical issues to possibly present PLT aggregates in a
transfused PLT product retrospectively.
The circulation and distribution of ADP-induced PLT
aggregates in vivo were studied by Valeri et al. [56] who
transfused 111-In-oxine-labelled autologous PLT aggre-
gates without a filter in baboons. Interestingly, the
labelled PLT aggregates were initially sequestered in the
lungs and then released into the peripheral blood within
the next 3 h. The clinical relevance of these findings,
however, and their relevance for currently performed
transfusion practice using standard filter routinely remain
unclear. The two anesthetized baboons being transfused
with the PLT aggregates showed normal blood pressure,
heart rate, respiration rate and temperature.
© 2013 International Society of Blood Transfusion
Vox Sanguinis (2014) 106, 209–218
Aggregates in platelet products 215
(b)
Managing and preventing PLT aggregation in PLT
products
As discussed above, many factors may play a role in the
formation of aggregates in PLT products during and after
preparation. As a result, definitive root causes can some-
times remain undetermined. However, some general sug-
gestions may promote the dissipation of aggregates when
they do occur. The addition of inhibitors of PLT activity
such as prostaglandin E1 to PRP-derived PLT products was
investigated in a small study. The results suggested less PLT
aggregate formation during storage as well as easier and
faster PLT resuspension after centrifugation. However, due
to the measured reduction in in vivo survival in this study,
this concept was not further pursued in routine production
[57]. If visible aggregates are present after the collection,
it could be beneficial to rest the PLTs at least for 1 h with-
out agitation, with the bag label-side down followed by
subsequent agitation for at least another hour (likewise
with the bag label-side down).
Environmental factors are often major contributors to
the formation of persistent PLT aggregates. They can both
inhibit the dispersion of and cause persistent aggregates.
As discussed above, lower temperatures tend to increase
PLT aggregation with the optimal storage temperature for
PLTs being 20–24°C. Exposing PLT product to tempera-
tures outside this range or placing products on cold sur-
faces should be avoided throughout the entire production
216 J. Ringwald et al.

and storage process. Furthermore, to ensure proper gas

exchange, PLT products should be stored flat and label-
side down.
Finally, donor-dependent variables may contribute to
persistent aggregation. If, for a specific donor, resting and
agitation do not dissipate aggregates, it would be interest-
ing to track the donor to determine whether the aggre-
gates may be donor specific. If PLT products donated by
a specific donor appear to consistently develop aggre-
gates, it should be considered to increase the AC delivered
during the apheresis procedure for this donor. Modern
apheresis technology offers this option, and typical values
for this configurable parameter can be found in various
publications [58,59]. As blood is drawn from the donor
into the system, AC and WB are mixed at the configured
inlet to AC ratio. The inlet to AC ratio expresses numeri-
cally how many parts of WB are mixed per part of AC.
Lowering the ratio will increase the amount of AC in the
extracorporeal circuit and eventually in the PLT product.
This can help prevent the formation of aggregates as pre-
viously discussed by lowering the pH gently. A recent
publication has reported positive result of lowering the
inlet to AC ratio to reduce the frequency of occurrence of
PLT aggregates [60].
Summary and conclusions
The occurrence of reversible and irreversible micro- and
macroaggregates is a phenomenon observed in PLT con-
centrates derived of WB or collected by apheresis. It is
most likely a multifactorial phenomenon of which the
root cause cannot always be determined. Overall, pub-
lished data on this topic remain scant. The published
data on in vitro test parameters comparing products with
and without aggregates suggest that the impact of a
small number of residual aggregates in PLT products
appears to be of limited impact on PLT in vitro quality.
No clear conclusions can be drawn with regard to the
clinical relevance of aggregates in PLT products as reli-
able data are lacking. The fact that any visible aggre-
gates are captured by a filter in the transfusion set
reduces the potential risk to the patient. Nonetheless,
considering the concern of clinicians about this phenom-
enon, more studies are needed which should focus on
the possible clinical impact of such aggregates and possi-
ble precautions to avoid PLT aggregate formation in PLT
products.

References
1 Mourad N: A simple method for
obtaining platelet concentrates free of
aggregates. Transfusion 1968; 8:48
2 Ringwald J, Eckstein R, Antoon M,
et al.: Occurrence of residual aggre-
gates in single donor platelet con-
centrates in routine blood banking. Vox
Sang 2013; 105(Suppl. 1):142–143
3 Gkoumassi E, Klein-Bosgoed C, Dijk-
stra-Tiekstra MJ, et al.: Noninvasive
pH monitoring of platelet concen-
trates: a large field test. Transfusion
2013; 53:2287–2292
4 Flatow FA, Freireich EJ: The increased
effectiveness of platelet concentrates
prepared in acidified plasma. Blood
1966; 27:449–459
5 Murphy S, Gardner FH: Platelet stor-
age at 22 degrees C: role of gas trans-
port across plastic containers in
maintenance of viability. Blood 1975;
46:209–218
6 Murphy S, Sayar SN, Gardner FH:
Storage of platelet concentrates at
22°C. Blood 1970; 35:549–557
7 Jaeremo P, Shaba A, Kutti J: Some
storage characteristics of buffy coats
used for preparation of platelet con-
centrates. Ann Hematol 1992; 65:269–
273
8 Skripchenko A, Myrup A, Awatefe H,
et al.: A rest period before agitation
may improve some in vitro apheresis
platelet parameters during storage.
Transfusion 2012; 52:1433–1438
9 Flesch BK, Adamzik I, Steppat D,
et al.: Paired crossover study of two
plateletpheresis systems concerning
platelet product quality and donor
comfort. Transfusion 2010; 50:894–
901
10 Welch M, Champion AB: The effect
of temperature and mode of agitation
on the resuspension of platelets
during preparation of platelet con-
centrates. Transfusion 1985; 25:283–
285
11 Sowemimo-Coker SO, Zinn F, Kim A:
Formation of macroaggregates and
clumps in platelet concentrates: effects
of processing conditions. Transfusion
2003; 43(Suppl.):29A
12 Devine D, Bradley AJ, Maurer E, et al.:
Effects of prestorage white cell reduc-
tion on platelet aggregate formation
and the activation state of platelets
and plasma enzyme systems. Transfu-
sion 1999; 39:724–734
13 Murphy S, Martinez J, Holburn R:
Stability of plasma fibrinogen dur-
ing storage of platelet concentrates
at 22°C. Transfusion 1983; 23:
480–483
14 Racz Z: Clumps present in platelet
concentrates are formed from fibro-
nectin. Vox Sang 1996; 71:242
15 Racz Z, Baroti C: Appearance of
clumps in stored platelet concentrates
due to intensive agitation. Vox Sang
1995; 68:249–250
16 Moroff G, Kline L, Dabay M, et al.:
Reevaluation of the resting time period
when preparing whole blood-derived
platelet concentrates with the platelet-
rich plasma method. Transfusion 2006;
46:572–577
17 Aster RH: The anticoagulants of choice
for platelet transfusions. Transfusion
1966; 6:32–38
18 Snyder EL, Ferri P, Brown R, et al.:
Evaluation of flatbed reciprocal
motion agitators for resuspension of
stored platelet concentrates. Vox Sang
1985; 48:269–275

© 2013 International Society of Blood Transfusion

Vox Sanguinis (2014) 106, 209–218

19 Anstall HB, Hawkey CM: Observa-
tions on platelet clumping and
related phenomena. Transfusion 1962;
2:44–51
20 Pert JH, Zucker MB, Lundberg A, et al.:
Recent advances in the preparation of
platelet concentrates from ACD and CPD
blood. Vox Sang 1967; 13:119
21 Kattlove H, Alexander B: The effect of
cold on platelets. I. Cold-induced
platelet aggregation. Blood 1971;
38:39–48
22 World Health Organization: Model
Standard Operating Procedures for
Blood Transfusion Service. E-Publica-
tion. New Delhi, India, World Health
Organization, Regional Office, 2002
23 AABB. Chapter 6: Blood collection,
storage, and component preparation:
Technical Manual 15th Edn. Bethesda,
American Association of Blood Banks,
2005:816
24 Wong T, Pedvis L, Frojmovic M: Plate-
let size affects both micro- and macro-
aggregation: contributions of platelet
number, volume fraction and cell sur-
face. Thromb Haemost 1989; 62:733–
741
25 Paulus JM: Platelet size in man. Blood
1975; 46:321–336
26 Bang NU, Heidenreich RO, Trygstad
CW: Plasma protein requirements for
human platelet aggregation. Ann N Y
Acad Sci 1972; 201:280–299
27 Jorgensen KA, Stoffersen E: On the
inhibitory effect of albumin on platelet
aggregation. Thromb Res 1980; 17:13–
18
28 Lam FW, Cruz MA, Leung HC, et al.:
Histone induced aggregation is inhib-
ited by normal albumin. Thromb Res
2013; 132:69–76.
29 Ringwald J, Zimmermann R, Eckstein
R: The new generation of platelet
additive solution for storage at 22
degrees C: development and current
experience. Transfus Med Rev 2006;
20:158–164
30 Gawaz M, Ott I, Reiniger AJ, et al.:
Effects of magnesium on platelet aggre-
gation and adhesion. Magnesium mod-
ulates surface expression of
glycoproteins on platelets in vitro and
ex vivo. Thromb Haemost 1994;
72:912–918
31 Weis-Fogh US: The effect of citrate,
calcium, and magnesium ions on the
© 2013 International Society of Blood Transfusion
Vox Sanguinis (2014) 106, 209–218
potassium movement across the
human platelet membrane. Transfusion
1985; 25:339–342
32 Ravn HB, Vissinger H, Kristensen SD,
et al.: Magnesium inhibits platelet
activity: an in vitro study. Thromb
Haemost 1996; 76:88–93
33 De Wildt-Eggen J, Schrijver JG, Bins
M, et al.: Storage of platelets in addi-
tive solutions: effects of magnesium
and/or potassium. Transfusion 2002;
42:76–80
34 Gulliksson H, AuBuchon JP, Vesterinen
M, et al.: Storage of platelets in additive
solutions: a pilot in vitro study of the
effects of potassium and magnesium.
Vox Sang 2002; 82:131–136
35 Gulliksson H, AuBuchon JP, Cardigan
R, et al.: Storage of platelets in addi-
tive solutions: a multicentre study of
the in vitro effects of potassium and
magnesium. Vox Sang 2003; 85:199–
205
36 Ringwald J, Walz S, Zimmermann R,
et al.: Hyperconcentrated platelets
stored in additive solution: aspects on
productivity and in vitro quality. Vox
Sang 2005; 89:11–18
37 Ashford P, Gulliksson H, Georgsen J,
et al.: Standard terminology for plate-
let additive solutions. Vox Sang 2010;
98:577–578
38 Garner SF, Jones CI, Stephens J, et al.:
Apheresis donors and platelet func-
tion: inherent platelet responsiveness
influences platelet quality. Transfusion
2008; 48:673–680
39 Vasconcelos E, Figueiredo AC, Seghat-
chian J: Quality of platelet concen-
trates derived by platelet rich plasma,
buffy coat and Apheresis. Transfus
Apher Sci 2003; 29:13–16
40 Metcalfe P, Williamson LM, Reuteling-
sperger CPM, et al.: Activation during
preparation of therapeutic platelets
affects deterioration during storage: a
comparative flow cytometric study of
different production methods. Br J
Haematol 1997; 98:86–95
41 Fijnheer R, Pietersz RN, de Korte D,
et al.: Platelet activation during prepa-
ration of platelet concentrates: a com-
parison of the platelet-rich plasma and
the buffy coat methods. Transfusion
1990; 30:634–638
42 Nakajo S, Hirayama F, Niwa K, et al.:
Clump formation in apheresis platelet
Aggregates in platelet products 217
concentrates. Transfusion 1999;
39:913–914
43 Hagberg IA, Akkok CA, Lyberg T, et al.:
Apheresis-induced platelet activation:
comparison of three types of cell sepa-
rators. Transfusion 2000; 40:182–192
44 Moh-Klaren J, Berriot N, Lecrubier C,
et al.: Platelet collection by apheresis
with reduced use of citrate: quantita-
tive and qualitative evaluation of three
protocols. Transfus Clin Biol 1995;
2:181–188
45 Nussbaumer W, Plattner R, Schoppel
W, et al.: PLT-30: an improvement for
platelet on the Baxter CS-3000 Plus. J
Clin Apher 1994; 9:236–239
46 Ringwald J, Duerler T, Frankow O,
et al.: Collection of hyperconcentrated
platelets with Trima Accel. Vox Sang
2006; 90:92–96
47 Sandgren P, Saeed K: Storage of
buffy-coat-derived platelets in additive
solution: in vitro effects on platelets of
the air bubbles and foam included in
the final unit. Blood Transfus 2011;
9:182–188
48 Boomgaard MN, Gouwerok CW, Palfe-
nier CH, et al.: Pooled platelet concen-
trates prepared by the platelet-rich
plasma methods and filtered with three
different filters and stored for 8 days.
Vox Sang 1995; 68:82–89
49 Sweeney JD, Dupuis M, Testa G: Lack
of macroscopic aggregates in pools of
prestorage WBC-reduced random-
donor platelet concentrates prepared
from whole blood. Transfusion 2000;
40:615
50 Pedigo M, Wun T, Paglieroni T:
Removal or white cell-reduction filters
of activated platelets expressing CD62.
Transfusion 1993; 339:30–35
51 Jimenez TM, Patel SB, Pineda AA,
et al.: Factors that influence platelet
recovery after transfusion: resolving
donor quality from ABO compatibility.
Transfusion 2003; 43:328–334
52 Sweeney J, Kouttab N, Holme S, et al.:
In vitro evaluation of prestorage pools
consisting of mixed A and O platelet
concentrates. Transfusion 2007;
47:1154–1161
53 Greco-Stewart VS, Brown EE, Parr C,
et al.: Serratia marcescens strains
implicated in adverse transfusion reac-
tions form biofilms in platelet con-
centrates and demonstrate reduced
218 J. Ringwald et al.
detection by automated culture. Vox
Sang 2012; 102:212–220
54 Greco CA, Maurer-Spurej E, Scott MD:
PEGylation prevents bacteria-induced
platelet activation and biofilm forma-
tion in platelet concentrates. Vox Sang
2011; 100:336–339
55 Feys B, Devloo R, De Pourcq K, et al.:
Apheresis platelet concentrates con-
taining visible particles are not inferior
to particle-free concentrates, in vitro.
Vox Sang 2013; 105(Suppl. 1):11
56 Valeri CR, Macgregor H, Giorgio A,
et al.: Circulation and distribution of
111-In-oxine-labeled autologous
baboon platelet aggregates and buffy
coat. Transfus Apher Sci 2005;
32:139–146
57 Hawker RJ, Turner VS, Mitchell SG:
Use of prostaglandin E1 during prepa-
ration of platelet concentrates. Trans-
fus Med 1996; 6:249–254
58 Bueno JL, Garcıa F, Castro E: A ran-
domized crossover trial comparing
© 2013 International Society of Blood Transfusion
three plateletpheresis machines. Trans-
fusion 2005; 45:1373–1381
59 Picker SM, Radojska SM, Gathof BS:
Prospective comparison of high-dose
plateletpheresis with the latest aphere-
sis systems on the same donors. Trans-
fusion 2006; 46:1601–1608
60 Douglas C, Cardenas C, Mills AE: Risk
vs. reward: setting the anticoagulant
ratio to maximize platelet collections
and mitigate aggregates. Transfusion
2012; 52:AP10
Vox Sanguinis (2014) 106, 209–218