Original Article
Standardized Counting of Circulating Platelet
Microparticles Using Currently Available Flow
Cytometers and Scatter-Based Triggering:
Forward or Side Scatter?
P. Poncelet,1* S. Robert,2 T. Bouriche,1 J. Bez,1 R. Lacroix,2,3 F. Dignat-George2,3
 Abstract
1
R&T Department, BioCytex, Marseille,
France
2
Faculte De Pharmacie, VRCM,
UMR-S1076, Aix-Marseille Universit
e,
INSERM, Marseille, France
3
Hematology and Vascular Biology
Department, CHU La Conception, APHM,
Marseille, France
Received 30 October 2014; Revised 23
March 2015; Accepted 15 April 2015
Additional Supporting Information may be
found in the online version of this article.
*Correspondence to: Philippe Poncelet,
Director Research & Technology
BioCytex 140, Chemin de l’Arm ee
d’Afrique 13010, Marseille, France.
E-mail: philippe.poncelet@biocytex.fr
Published online 11 May 2015 in Wiley
Online Library (wileyonlinelibrary.com)
DOI: 10.1002/cyto.a.22685
C 2015 International Society for
 Key terms
V
Advancement of Cytometry
Cytometry Part A  89A: 148 158, 2016
Clinical determination of MP counts using flow cytometry has not been fully accepted
yet due to the lack of standardization protocols. In the past 5 years, we have proposed
two versions of a method with reproducible PMP counts in plasma samples. Both
methods use forward scatter (FSC)–based threshold set with reference beads of appro-
priate sizes; first using 0.5 mm beads and later with 0.3 mm beads. Both systems provide
reproducible PMP counts. However, this technique works only with some of currently
used commercial flow cytometers. Instruments with limited resolution or generating
heterogeneous FSC signals are excluded. Such performances are incompatible with the
required interinstrument standardization. Here we show that (i) flow cytometers with
sub-optimal FSC capabilities generally have higher SSC resolution and background
rejection capacity, and (ii) that the same biological entities, “dim and bright PMP,”
both can be counted using alternative strategies, either as previously described, based
on FSC measurements, or as presented here, based on SSC detection. The critical ele-
ment in the standardization protocol is the use of different sizes of reference beads.
This study was designed to permit simultaneous access to both FSC- and SSC-
optimized platforms. A new range of about 0.17–0.6 mm eq. (mm-equivalents) is pro-
posed for an alternative SSC–based MP gate generating the same PMP counts as those
obtained in the previously proposed 0.3–1 mm eq. FSC–based MP gate. The two equiv-
alent standardization options reconcile intrinsically different scattering behaviors
between SSC- and FCS - triggered instruments and open the opportunity for multicen-
ter studies in the future. VC 2015 International Society for Advancement of Cytometry
absolute counting; microparticles; microvesicles; standardization; submicron particles;
extra-cellular vesicles; scatter; interplatform reproducibility
MICROPARTICLES (MP) are sub-micron sized vesicles released from cell mem-
branes in response to activation or apoptosis. They are generally defined as 0.1–1 lm
closed membrane fragments exposing the anionic phospholipid phosphatidylserine
(PS) and membrane antigens representative of their cellular origin. MP originating
from several cell sources have been described in human plasma. Among them,
platelet-derived MP (PMP) are believed to account for the majority of circulating
MPs in healthy subjects (1). In clinical practice, circulating MP originating from
blood and vascular cells are elevated in a variety of prothrombotic and inflammatory
states, cardiovascular or autoimmune disorders and infectious diseases, as well as
malignancies (2,3). Although MP counts may provide useful diagnostic/prognostic
information, assessment of their pathophysiological relevance in multicenter studies
is hampered by a lack of consolidated method and of standardized protocol (4).
Flow cytometry (FCM) is the most commonly used technique among methodologies

available to measure MP in biological samples (5). Many
attempts have been made to overcome resolution limitations
to bring the power of FCM down-to the nanoparticle sizing.
Recently, some astonishing success has been reported (6).
Such breakthrough that allows the measurement of all extra-
cellular vesicles, including all MP and possibly exosomes, is
improvement of utmost interest involving new technology. In
our study however, the objective is to support improved pro-
tocols which permit the use of as many currently available
flow cytometers (FCs) as possible for standardized MP counts.
This objective is a significant challenge as we are facing an ice-
berg with only a small detectable visible tip. The largest part
of the elements of interest are still undetectable. However,
powerful clinical information can already be deciphered from
detection of the visible part (7–9). Five years ago, we proposed
a simple strategy based on size-calibrated latex beads allowing
several labs to set-up reproducible forward scatter (FSC)
conditions including a well-defined common window of
analysis (10). This strategy was applied in an international
workshop where several laboratories worldwide obtained
reproducible PMP counts. The successful instruments were
mainly Beckman-Coulter (BC) FC500 or Epics XL cytometers
(11). However, with a large group of FCs including those
from Becton-Dickinson (BD), the FSC-based strategy was not
perfectly working and interplatform discrepancies in PMP
counts appeared. A rather heterogeneous individual instru-
ment behavior was observed both in terms of FSC resolution
and positioning of PMP in the standardized FSC scale. Intrin-
sic technological differences in FSC collection are responsible
for these discrepancies, mainly due to using different solid
angles to collect FSC light signals. Flow cytometers with low
FSC collection angle may well counter-balance this intrinsic
limitation of FSC resolution by a good resolution and back-
ground rejection capacity using side scatter (SSC), as already
suggested (12–14). The aim of this study thus is (i) to demon-
strate the superior resolution of SSC parameter when using
FSC-limited instruments in the sub-micron range, (ii) to
define which reference beads may be used for the SSC scale,
and (iii) to provide an alternative strategy for future interplat-
form standardization of PMP counts and possibly all MP.
MATERIALS AND METHODS
Instruments
The reference instrument was a standard three laser BC
Gallios used for MP analysis and regularly checked using
Megamix-Plus FSC as a quality control tool (see Supporting
Information Fig. 9 for details). Thanks to a special option to
collect FSC at wider angles than classically done (signals at 98–
198 only are amplified whereas signals at 18–98 are not), this
type of FC allows for superior resolution of submicron par-
ticles. This permits systematic visualization of 0.1 mm fluores-
cent beads when triggering with FITC (FL1) fluorescence, as
illustrated in Supporting Information Figures 1 and 2 (15). As
part of our regular MP-oriented quality-control (Q.C.) pro-
cess, background level was determined using 0.22 mm filtered
water run as a sample in the standardized 0.3 mm eq. FSC cut-
Cytometry Part A  89A: 148 158, 2016
Original Article
off conditions. It was lower than 2,000 events/s, which was
considered acceptable with respect to the data treatment
capacity of the electronics (max acquisition rate 25,000
events/s, see Supporting Information Fig. 9 and Supporting
Information Table 2 for details). The scatter distribution of
submicron beads cocktails was also tested on various instru-
ments including several BC Gallios/Navios, BD FACS-Canto,
LSRII, FACS-Verse, Fortessa, and Stratedigm S-1000 (see Sup-
porting Information Table 1 for listing and correspondence
with figure numbers).
Aside from the above cited BC Gallios, a BD Fortessa
with small particle detection unit was used to compare MP
versus beads behavior in both FSC and SSC parameters on a
single instrument. This system has recently been described
(16). It uses a special focalization optics with internal laser
blocking mask and a photomultiplier tube. It still leaves access
to the standard diode-supported FSC collection module
through the use of a 90/10 beam splitter, thus allowing two
measurements in parallel, one standard FSC plus one addi-
tional FSC parameter noted PMT-FSC. Although Rousseau
et al. (16) suggested that their small particle module enlarged
the scatter collection solid angle up-to 28.58, this value is
highly dependent on the precise position, shape, and dimen-
sion of the blocking mask (or bar) and may vary between
instruments. The small particle detection module of our BD
Fortessa had been previously optimized for maximal resolu-
tion by BD engineers using internal procedures. Without any
specific information, we used our beads with the preset optical
configuration. A favorable resolution in the FSC-PMT scale
was indeed observed (Fig. 3), allowing the use of FSC-PMT in
addition to SSC and the standard FSC. The 488 nm blue laser
was run at 50 mW power while no other laser was used in the
present study.
Sample Collection and Preparation
Healthy staff members donated citrated blood samples
with informed consent. Platelet-free plasma (PFP) was pre-
pared and frozen at 2808C as previously described (17). To
generate various levels of PMP, some blood tubes were agi-
tated from 10 to 60 min by head-over-head mixing at room
temperature (RT) on a rotating wheel. In a few cases, MP-free
plasma (MPFP) was used to dilute some aliquots of PFP after
thawing to generate very low levels. MPFP was created from
normal pooled human plasma by two successive steps of
ultra-centrifugation for 90 min at 75,000g. Finally, several dif-
ferent crossed associations of PFP were also created, thus gen-
erating a total of 28 PFP samples featuring different PMP
counts.
Reagents and Sample Staining
For direct comparison, each stained sample was split after
final dilution into appropriate tubes (4 ml polystyrene FACS
tubes, Falcon, Becton-Dickinson, Grenoble, France for the BD
Fortessa and 4 ml polypropylene FCM tubes, Dutscher, Stras-
bourg, France for BC Gallios). All duplicate tubes were kept at
RT in the dark for a maximum of 1 h, to avoid accidental clot-
ting, and were run in parallel on both instruments. Staining
was performed as previously described (15) with minor
149
Original Article
changes. These included doubling all volumes for final split-
ting prior to parallel FCM analysis. Briefly, 60 ml of PFP were
incubated with 10 ml of CD41-PE (BioCytex, Marseille,
France) and 20 ml Annexin V-FITC diluted 1/30 (Tau Technol-
ogies bv, Kattendijke, NL) in its paired Ca21 containing Bind-
ing Buffer for 30 min at RT in the dark. Following final
dilution with 1 ml of Binding Buffer, the counting beads were
R
added (60 ml; Cyto-CountV beads, DAKO, Trappes, France).
Just prior to analysis, each tube was homogenized by stirring
with a vortex, either manually (Fortessa) or automatically
(Gallios).
Instrument Settings and Data Management
R
Filtered sheath fluids (0.22 mm) (FACS-FlowV in BD For-
VR
tessa; Isoflow in BC Gallios) were used. To regularly clean
biological deposits in tubing and internal walls of the flow-
R
cells, fluidic lines were filled every night with BC ClenzV
cleaning solution and thoroughly rinsed every morning with
filtered sheath fluid. Three versions of the standardization
beads (BioCytex) were used according to instruments i.e.,
Megamix-Plus FSC (for use on FSC-optimized platforms),
Megamix-Plus SSC (for use on SSC-optimized platforms) and
also a v/v mixture of both, called Gigamix, for instruments
with appropriate resolution in both scatter parameters.
Instruments were set-up and controlled prior to the analysis
of any new MP series using Megamix-Plus beads according to
manufacturer’s instructions. As it will be summarized in Sup-
porting Information Fig. 9 to conclude the present study,
these include a first step of fluorescence trigger to visualize all
beads, check resolution, and set-up scatter and FITC fluores-
cence conditions. The second step requires switching to scatter
triggering with threshold level standardized with the help of
reference beads.
In addition to official directives, in the Gallios, PEAK sig-
nals were also selected to be included into the raw data for all
parameters in addition to integral signals (INT). From previ-
ous experience with BD instruments it was demonstrated that
amplitude signals (-H) generally provide better resolution
than composite area signals (-A), as previously illustrated
(11). This is also illustrated in Supporting Information Figure
1 for two representative cytometers, a BD FACS-Canto II and
a BC Gallios. In addition, this option also provides a direct
visualization of any threshold since triggering is based on sig-
nal amplitude and not area. A standardized 0.3 mm eq. FSC-
based threshold was adapted on the Gallios using the 2:1
numerical ratio between 0.3 and 0.5 mm beads as described
(15). On the Fortessa, a very open SSC-based threshold was
selected in order (i) to fit with BD instruments’ preferential
use of SSC as a trigger parameter and (ii) to restrict the analy-
sis only during retreatment of acquired raw data. This was
possible because of an exceptionally low background noise in
this individual instrument. Fluorescent beads (FITC-like 0.5
mm beads from Megamix and 3 mm PE-coated beads (BioCy-
tex)), with intensity levels in the same range as bright PMP,
were used as target channel systems in order to position PMP
at similar levels in the dual fluorescence plots in each FC. For
each sample, acquisition was stopped after 1 min and
150
list-mode files automatically stored as fcs 3.0 compatible files.
Re-analysis of data files was done using KaluzaTM (BC, Ville-
pinte, France) as detailed in the results section. It is notewor-
thy that KaluzaTM software represents the last four decades
from raw data acquired with respectively 1 million and
256,000 channels dynamic range in Gallios and Fortessa and
display the data on a 0.1–1,000 a.u. logarithmic scale (vs. 1–
10,000 a.u.–alternatively 256,000 a.u.—for BD DIVA soft-
ware). These differences were considered to define common
target channels for fluorescence as well as for SSC when
needed as in the case of Figure 2. Linear regression analysis
R
was performed using Microsoft ExcelV.
RESULTS
Scatter Heterogeneity Among Various Types of Flow
Cytometers
Scatter performance in the sub-micron field varies
among instruments according to the scatter parameter used.
BC Gallios/Navios display high resolution in FSC, as illus-
trated in Supporting Information Figure 1, with rather homo-
geneous interinstrument characteristics. A few other types of
FCs also showed favorable FSC resolution as illustrated in
Supporting Information Figure 2.
In contrast, many other types of FCs showed heterogene-
ous and most often low FSC resolution, as demonstrated with
8 different instruments (Fig. 1 and Supporting Information
Fig. 3). Only the biggest 0.5 and 0.9 mm beads from Megamix-
Plus FSC were shown as FSC histograms since in most of these
cases (7/8), smaller beads could not be resolved from the
0.5 mm beads. On the contrary, all 4 beads were nicely discri-
minated using SSC and the relative position of each bead
subset was reproducible among the 8 instruments. These
observations demonstrated the superior and homogeneous
resolution of SSC parameter in this group of FSC-limited
instruments. We called the FCS and SSC groups “FSC-
optimized” and “SSC-optimized” platforms, respectively. The
SSC group was characterized by a reproducible wide gap
observed in SSC between 0.1 and 0.3 mm beads (Fig. 1 and
Supporting Information Fig. 3). Three additional fluorescent
beads of 160, 200, and 240 nm were selected to cover the
dynamic SSC scale together with the standard Megamix 0.5
mm beads, thus creating a new blend called Megamix-Plus
SSC. The new beads did not overlap with each other nor with
the already existing Megamix-Plus FSC beads. All seven differ-
ent sizes covering a range of 100–900 nm could be simultane-
ously analyzed using a simple v/v mixture of both Megamix-
Plus blends, called Gigamix. This was used to illustrate the
resolution of multiple SSC-optimized FCs (Supporting Infor-
mation Fig. 4). Supporting Information Table 1 provides a
tentative classification of many commercial FCs that we had
the opportunity to test.
Intensity of SSC is differentially associated with bead
diameter in the 300–900 nm and in the 100–300 nm size
ranges based on the use of the 7 Gigamix beads (Fig. 2). Below
300 nm diameter the decrease is with a much steeper variation
in SSC intensity. Interestingly, when SSC gain was harmonized
using 0.5 mm beads to target the same location in the SSC
Standardization of Microparticle Counts
 Original Article

Figure 1. High and reproducible scatter resolution among sub-micron beads depends on scatter parameter in a sub-group of commercial
flow cytometers. Megamix-Plus FSC beads were run in various BD instruments under ideal conditions (selection of amplitude signals and
low flow rate). FSC distribution between the two largest submicron beads only (0.5 and 0.9 mm, same as in standard Megamix) is illus-
trated in A–D. The corresponding SSC distributions of all four submicron beads are shown in the lower row (E–H). FC models include
FACS-Verse (A, E), Fortessa (B, F), and two individual FACS-Canto II (C, G, and D, H). 4 additional examples can be viewed in Supporting
Information Figure 2.

scale (4 upper lines), the peak positions of all beads were
almost superimposable between instruments from different
generations, e.g., FACS-Canto II versus FACS-Verse and For-
tessa, but also from different manufacturers e.g., BD versus
Stratedigm. Even when the beads distribution was voluntarily
located at a different place in the SSC scale (lower line) the rel-
ative bead positions remained comparable, thus providing a
favorable basis for future standardization. In summary, flow
cytometers with sub-optimal FSC capabilities generally have
higher SSC resolution and more homogeneous response with
submicron beads. Two differentiated groups of FCs should
thus be considered for interplatform standardization. As the
next step of this study, we had to select within the laboratory
one representative instrument of each sub-group. Parallel
analysis of the same samples became possible using a BD
Fortessa installed alongside a BC Gallios.

Characteristics of Scatter Standardization Beads in

Figure 2. Homogeneous distribution of sub-micron beads in vari-
ous SSC-optimized instruments. Megamix-plus FSC beads
enriched with 160, 200, and 240 nm beads (referred to as Giga-
mix) were run in various SSC-optimized instruments either (i)
similarly set-up using a common position of the 500 nm beads in
the SSC scale (target channel equivalent to 300 a.u. in the present
0.1–1,000 a.u., 4 decade scale): upper traces for FACS-Canto II,
FACS-Verse, Fortessa, and Stratedigm A or (ii) set-up differently
with a lower position of the 500 nm beads (equivalent to 60 a.u.):
lower trace for Stratedigm B. Median values for each bead subset
in each instrument are plotted against nominal bead diameter.
Best-fit exponential-type tendency curves (dotted lines) are drawn
independently (i) for each SSC setting and (ii) for two groups of
bead sizes i.e., from 100 to 300 nm on the left side and from 300
to 900 nm on the right side of the graph.1Gigamix 5 v/v mixture
of both Megamix FSC and SSC, see Methods section.
Various Platforms
To set-up both FCM systems with the most useful scatter
reference beads for standardization purposes, scatter profiles of
the selected beads were characterized on each platform. For an
FSC-optimized platform such as the BC Gallios, we used
Megamix-Plus FSC beads as already published (15). This blend
of beads covers a wide range from 100 nm up to 1,000 nm
(Fig. 3.1 A–D). According to manufacturer’s instructions, initial
selection of bead subsets (100, 300, 500, 900 nm) is made, using
fluorescence threshold, on the FL1 vs. SS density plot showing
the characteristic positions of each of the four major clouds of
singlets (Fig. 3.1.A). It is noteworthy, that the 300 nm beads har-
bor a higher FL1 intensity than the 500 nm beads, thus provid-
ing a specific profile and avoiding significant contamination of
500 nm singlets with 300 nm doublets. Also, the 100 nm beads

Cytometry Part A  89A: 148 158, 2016 151

Original Article

Figure 3. FCM profiles of scatter standardization beads in various platforms. The standardization bead cocktails used in this study are
depicted as they appear on the three different platforms after fluorescence (FL1/FITC) triggering. 1) Megamix-Plus FSC beads on the BC
Gallios and 2) on the Fortessa with PMT-based FSC detection. 3) Megamix-Plus SSC beads on the Fortessa using standard diode-based
FSC detection. 4) Gigamix on the Fortessa with PMT-based FSC detection. In each case, selection of singlets in each bead subset is oper-
ated on the SSC vs. FITC density plot A. Color dot-plot D shows the dual scatter distribution of all singlets. Plots B and C show SSC and
FSC histograms of all gated beads, respectively. Amplitude signals (PEAK or –H) are used throughout.

clearly appear as a discriminated cloud. Superior resolution
using FSC versus SSC is obvious in Fig. 3.1.B–3.1.D. Interest-
ingly, the Fortessa equipped with small particle detection system
provided a unique opportunity to compare the relative positions
of these beads on both scatter scales. As illustrated in Fig. 3.2,
not only the discrimination of 0.1 versus 0.3 mm beads was sig-
nificant in SSC (Fig. 3.2.B), as expected from previous data on
similar machines (e.g., Fig. 1), but an almost complete resolu-
tion was also observed in FSC (Fig. 3.2.C), as with Gallios. The
same Fortessa used with the diode-based FSC detector was
clearly not usable as an FSC-optimized instrument since 0.5 mm
beads were hardly resolved from smaller beads (e.g., 0.24 mm,
Fig. 3.3.C). On the contrary, the optimal SSC resolution permit-
ted a clear discrimination of all beads from the Megamix-Plus
SSC blend, including 160, 200, 240, and 500 nm (Fig. 3.3.B).
Thanks to this advantage and using these two instruments, we

152 Standardization of Microparticle Counts

 Original Article

Figure 4. Comparative interplatform PMP analysis. The same tube stained for PMP (CD41-PE/AnnV-FITC) was analyzed on two instru-
ments, Gallios 1) and Fortessa (2 and 3), using two different scatter-based standardization strategies (FSC in 1 and 2, SSC in 3). Trigger
was FSC for Gallios with threshold set at 0.3 mm eq (median of 0.3 mm beads), as shown in plot B. For Fortessa, trigger was SSC in both
cases with a purposely low threshold, which included most 0.16 mm beads, as depicted in plot K. Conditions of acquisition and MP gating
are summarized in the sub-table 3) and a letter « F » for « Fortessa » was added to region names in blocks 2 and 3. All parameters were
measured from their amplitude (Peak or Height) signals. Dual scatter color dot-plots A, E, and I show all events over each instrument’s
threshold in gray, except for the CD411/AnnV1 PMP. These are sub-divided into bright PMP (in blue) and dim PMP (in red), as also seen
in dual fluorescence plots C, G, and J, with the help of contour-plots (D and H). FSC-based gating regions (« (F-) Total MP »; « (F-) Large
MP », blocks 1 and 2) were located with the help of Megamix-Plus FSC beads shown in graphs B and F according to the following rules:
medians of 0.3 and 0.5 mm were used to set the lower and intermediate levels respectively and the end (99%-tile) of 0.9 mm beads to locate
the upper level. SSC-based gating regions (« L » for large MP; « T » for total MP) in graph I were located with the help of Megamix-Plus
SSC beads (0.16, 0.2, 0.24, and 0.5 mm, graph K), according to the following rules: (i) upper SSC level set at the end of 0.5 mm beads, (ii)
intermediate level set in the middle of 0.2 and 0.24 mm beads, (iii) lower level (Min value for region sT) between 0.16 and 0.2 mm beads,
precise value to be defined in the present study. Regions B, D, and homologs in scatter-gated plots C, G, and J provided the numbers of
PMP per run (1 min) that can be compared among platforms: e.g. [FB 1 FD] for total PMP to be compared with [FsB 1 FsD]) for both
platforms of the Fortessa. Contour plots D and H helped locating the two major PMP subsets in a similar manner on each instrument.
Although these were not gated by scatter-based MP gates as done for plots C, G and J, the overwhelming background events were
gated-out (« BkGd out») to generate helpful contours for PMP.

were able to compare three different MP analyzing platforms,
depending on the selection of the FSC parameter, FSC-H or
PMT-FSC-H, on the Fortessa. Although, each optimized
Megamix-Plus bead cocktail (SSC or FSC, respectively) could be
used individually, a convenient working option to get informa-
tion from the same run was to create a (v/v) mixture of both
(so-called “Gigamix”). Figure 3.4 shows the corresponding pro-
files and Plot B illustrates the resulting complete SSC histogram
with 7 discriminated peaks. Next steps will illustrate that only 4
out of those 7 beads are really useful for the calibration of the
SSC scale.
Definition of Threshold and PMP Gates in
Various Platforms
The PMP profiles of the same MP suspensions were com-
pared using different platforms to set-up the appropriate MP
gates (Fig. 4). First, scatter-based thresholds had to be defined.
To prepare the instruments for the acquisition of PMP, the
threshold was switched from fluorescence, as illustrated in Fig-
ure 3, to scatter, FSC or SSC, as shown in Figure 4. On the Gal-
lios, the FSC-based threshold was set-up in the middle of the 0.3
mm beads distribution, as already described (15) except that
only signal amplitudes (Peak/Height) were used as indicated in
the Methods section. Since in FCM, electronic threshold is
directly applied on amplitude signals, the lower half of the 0.3
mm beads’ FS PEAK distribution was precisely cut-out (Fig. 4B).
On the Fortessa, a very low SSC-based threshold could be
selected. Because of the low background on this instrument,
we could reproducibly open the analysis down-to a threshold
located below the 0.16 mm beads’ SSC distribution (Fig. 4K).
This low SSC threshold left all 0.3 mm, but not 0.1 mm beads,
to enter the analysis (Fig. 4F). In such conditions, the back-
ground detected in clean, 0.22 mm-filtered, distilled water was
always lower than 3,000 events per second, a rate that
remained far below the treatment capacity of the instrument’s
electronics (40,000 events/s, Supporting Information Table 2)
and thus kept the intrinsic coincidence rate to an acceptable
level.
Second, upper and lower limits of the MP gate were set
up. On the Gallios, the CD41/Annexin V staining of normal
PFP revealed two subsets of PMP (Figs. 4C and 4D), a bright
one and a dim one, as expected (15). Both subsets, colored in
blue and red respectively, grossly corresponded to large and
small MP (Fig. 4A), when the intermediate limit between both
subsets was defined in the FSC scale by the median of 0.5 mm
beads. The upper limit was provided by the end of the 0.9 mm

Cytometry Part A  89A: 148 158, 2016 153

Original Article
Figure 5. Comparison of PMP counts derived from various plat-
forms. PMP counts (expressed as counts per min) were compared
among different platforms. Graph A correlates PMP counts
between two platforms in a single instrument using two different
standardization strategies, based on FSC and SSC, respectively.
Graph B correlates PMP counts between two instruments using
the same, FSC-based, standardization strategy. Total PMP counts/
min in Fortessa were normalized to those in Gallios using
ungated bright PMP (regions « (F.) bright PMP » in Fig. 3D and
3H) as internal invariant reference. Graph C correlates PMP
counts between two instruments using two different standardiza-
tion strategies, based on FSC for the Gallios and on SSC for the
Fortessa, respectively.
beads and the lower limit corresponded to the previously
described 0.3 mm eq. threshold (Figs. 4A and 4B). In such
standardized conditions of analysis, the potential so-called
“swarming artefact” (16,18,19) has been ruled-out before con-
sidering the interplatform correlation study, as detailed in
Supporting Information Fig. 5.
154
On the Fortessa, the same two PMP subsets were also
observed and gated using the same contour-aided approach
than for Gallios (Figs. 4G and 4H). In the Fortessa FSC-
oriented platform, comparable limits could be applied in the
FSC scale (Figs. 4E and 4F), with the 0.5 mm beads also pro-
viding an appropriate intermediate limit between small and
large PMP. The same FSC-based MP gate (“F.Total MP” in
Fig. 4E) than that defined in the Gallios (“Total MP” in Fig.
4A) could be applied.
As stated in the summarized conditions sub-table of Fig-
ure 4, the Fortessa FSC-oriented platform required the use of
an SSC trigger since a FSC trigger at 0.3 mm eq. (as with Gal-
lios) had generated unacceptable background levels in prelimi-
nary experiments. Interestingly, the very low SSC threshold
applied during this widely open acquisition had rejected very
few of the small PMP that a standardized FSC-based threshold
would have accepted. Those would have been located over the
“F.Total MP” lower limit and left to the SSC threshold-induced
vertical cutoff, as shown by the arrow in Figure 4E. In the SSC-
oriented Fortessa platform (Figs. 4I, 4J, and 4K), the positions
of the SSC-based MP gates were significantly different from
those of FSC-based MP gates as compared to reference beads.
The upper limit of the large PMP subset could efficiently be
defined using the end the 0.5 mm beads, well below the 1
mm eq. upper limit previously used in FSC. The intermediate
limit between small and large PMP fitted well with the middle
of both 0.2 and 0.24 mm beads (minimum of the “sL” marker
in Fig. 4K), far below the 0.5 mm eq. level observed in FSC. The
lower limit of an SSC-based MP gate, that selects the same
amounts of PMP as the 0.3 mm eq. FSC-based MP gate, was
located somewhere between 0.16 and 0.2 mm beads and its
precise location had to be determined. This determination was
key to the process of FSC to SSC standardization.
SSC-Based Strategy for Interplatform
Standardization of PMP Counts
To define the new SSC-based standardization strategy,
PMP counts were compared first between the FSC- and SSC-
optimized platforms of the same instrument (Fortessa) to avoid
bias related to absolute counting. Next, the FSC-optimized plat-
forms of two different instruments (Fortessa FSC vs. Gallios
FSC) and finally the FSC-optimized Gallios platform and the
SSC-optimized platform of the Fortessa were compared.
First, we had to determine the optimal location of the
SSC-based MP gate. To achieve that, the minimum value in
SSC was modified step-by-step looking for the best correlation
among platforms. This was obtained by reanalyzing the elec-
tronic files (n 5 28), each time modifying the minimum level
of the SSC-based MP gate (“T” in graph I), calculating the
total PMP counts in scatter-gated plots (“F.D1F.B” in Fig. 4G
and “FsD 1 FsB” in Fig. 4J) and plotting total PMP per run
(PMP/min in Fig. 5) obtained with one platform against the
ones calculated using the other. When this SSC limit, corre-
sponding to the minimum value of the “sT” window (Fig. 4K)
was found, its location provided in arbitrary units of SSC was
related to the actual instrument settings and the type of log
scale used (e.g. 5.5 or 4 decades?, 0.1–1,000 a.u. or 1–10,000
Standardization of Microparticle Counts

Figure 6. Scatter scale standardization for MP analysis with vari-
ous platforms.This scheme is the authors’ personal view of the
problem debated in this study. The scatter distribution of plasma
PMP is here compared to a multipart floating body whose differ-
ent parts (subsets) are located by reference to water-emerging
vertical poles featuring both possible scales, FSC and SSC. The
emerging (and detectable) part of this floating body comprises all
the so-called « Large PMP » and a part of the whole « Small PMP
» subset. Theoretically, a 3rd (as yet totally undetectable) part
would include platelet-derived exosomes. In this scheme, the
whole « Small PMP » subset tentatively ranges between 0.1 and
0.5 mm eq. in the FSC scale (left). In the SSC scale (right), it begins
at an as yet undefined location to finish around 0.22 mm eq. In this
system, the water maximum height, including the waves featur-
ing instrument variations in background exclusion capacity,
stands at 0.3 mm eq. in the FSC scale and near 0.17 mm eq. in the
SSC scale. The totally visible « Large PMP » subset is between 0.5
and 1 mm eq. in the FSC scale, whereas it begins around 0.22
mm eq. to finish around 0.6 mm eq. in the SSC scale. The triangles
show the reference beads proposed here to help locate the vari-
ous upper-cited limits. This scheme based on our observations
on PMP may also apply to all types of MP, although their subsets
may not be divided the same way between small and large MP.
This explains why the title of Fig. 6 extends to MP, not only PMP.
a.u.?). To stay independent of the SSC scale settings, it was
decided to define this SSC limit by its distance relative to both
0.16 and 0.2 mm beads using the following formula:
Min “sT” (a.u.) 5 Md1 1 z* (Md2-Md1) where Md1 5
median of 0.16 mm beads, Md2 5 median of 0.2 mm beads and
0 < z < 1.
The best correlation was obtained with z 5 0.12 (Fig.
5A), with a slope very close to 1.0 (y 5 0.98x, r2 5 0.86). This
location at 12% of the distance between both beads was finally
selected as the one optimally relating both strategies on the
same instrument.
Next, we aimed to demonstrate that both FSC and SSC
strategies were applicable with two different instruments that
were not expected to have the same scatter scales, the same
PMP dual fluorescence aspects nor the same sample speed.
First, scatter scales were standardized as explained above. Sec-
ond, the analysis of PMP fluorescence staining was focused on
the two major subsets, optimally defined with the help of
contour-plots, as illustrated in Figures 4D and 4H. This
avoided the dual-labeled outlier events that were always pres-
ent in the dual positive quadrant (“PMP11” in Figs. 4C, 4G,
and 4J). Third, the sample speed normalization was initially
envisaged with the use of counting beads as already described
Cytometry Part A  89A: 148 158, 2016
Original Article
(10,15). For this purpose, known amounts of 5 mm Cyto-
CountTM counting beads were included into all stained sam-
ples, but determining the absolute PMP counts in the Fortessa
with the help of these counting beads turned-out to be impos-
sible. Although bead counts per run among different samples
were reproducible in the Gallios (mean 6 SD 5 442 6 16,
CV 5 4%, range: 411–462, first series of n 5 8, Supporting
Information Fig. 6A), the variation in the Fortessa was too
high (mean 6 SD 5 135 6 24, CV 5 18%, range: 96–168) to
allow comparison of absolute PMP counts among instru-
ments. Alternatively, with the same samples, the PMP counts
in the bright PMP subset were highly correlated between
instruments (y 5 2.68x; r2= 0.97, Supporting Information Fig.
6B). Thus, we decided to use the bright PMP subset as an
internal control for normalization. This was rendered possible,
in this particular case, by the use of the same suspensions in
all platforms at the same time.
As illustrated in Figure 5B, PMP counts were measured
on both instruments using the same FSC standardization
strategy and normalized to the sample speed of the Gallios as
explained above. They were similar (y 5 1.02x; r2 5 0.87)
demonstrating that when two different instruments display
similar distributions of sub-micron beads in the FSC scale, the
same FSC-based strategy can be applied to standardize PMP
counts among them.
Next, PMP counts measured on two different instru-
ments, each using its most appropriate scatter parameter and
standardized MP gate, were compared using the same sample
speed normalization approach (Fig. 5C). Similar counts were
also obtained in this case (y 5 1.03x; r2 5 0.90). Thus, this
three-platforms/two instruments study suggests that it is
possible to reach reproducible PMP counts on different
instruments provided that the described process is followed to
set-up scatter-based common windows of analysis between
FSC- and SSC-optimized instruments.
DISCUSSION
In this study, three different FC platforms have been
compared, looking for the standardization of scatter-based
settings on different instruments, using either forward scatter
(FSC) or side scatter (SSC) as the size-related parameter to
trigger acquisition and define a common window of analysis.
The rationale was to enumerate the same recognizable biologi-
cal objects (PMP) during parallel analysis of the same suspen-
sions on the different platforms, searching to equate PMP
counts across platforms and define the references to select in
order to obtain reproducible interplatform counts.
This study had several advantages. First, it profited from
the simultaneous evaluation of two instruments in a single
lab, offering options to analyze the same stained preparations
simultaneously. Second, we established conditions where
comparable settings were used not only for scatter parameters
but also for fluorescence. Third, artifacts associated with coin-
cidence, the “swarming effect,” were eliminated under the
conditions used by adjusting sample speed just high enough
to establish stable threshold conditions and to avoid the
155
Original Article
presence of significant doublet events. Fourth, the same soft-
ware was used for reanalysis of raw data. Fifth, selecting clearly
identified PMP subsets helped to reliably enumerate the same
biological objects on different platforms. Sixth, the unex-
pected although not uncommon difficulty encountered for
normalization of counts between cytometers with counting
beads could be overcome with the availability of an internal
invariant PMP subset.
In the absence of stabilized biological standards, we took
opportunity of a characteristic feature frequently observed in
plasma samples, the presence of two subsets of bright and dim
PMP that could be well discriminated. These subsets also cor-
respond approximately to high and low scatter intensity pop-
ulations of PMP, and thus were called “large PMP” and “small
PMP,” respectively (15). Discrimination of positive MP from
negatives using wide quadrants is prone to major uncertain-
ties during reanalysis and requires generating optimal negative
controls. Our proposed approach using tight elliptical regions
and contour-plots bypassed such difficulties. Having previ-
ously proposed a standardized FSC-based minimal cutoff at
the level of 0.3 mm eq. (15), we searched for an equivalent
SSC-based cutoff for other FCs with limited FSC but possibly
higher SSC resolution.
Preliminary investigations suggested that the position of
currently detectable PMP relative to reference polystyrene
beads may not be the same depending on the chosen size-
related parameter (P. Poncelet, Oral communication, CYTO
2012). This was demonstrated here using the same instrument
featuring two different platforms, each optimized for either
FSC or SSC. This was possible thanks to the dual capability of
a BD Fortessa with both resolution among 100, 300, 500, and
900 nm beads in FSC as well as resolution among 160, 200,
240, and 500 nm beads in SSC. Parallel analyses were operated
at the same time with replicates of the same stained samples,
thus eliminating all errors linked to (i) variability among ali-
quots of frozen plasma samples, (ii) staining related differen-
ces, and (iii) time delays from staining to FCM analysis.
The counting beads strategy used to derive absolute
counts of MP per microliter of plasma was not applicable with
statistical confidence on the Fortessa. In fact, this is a common
limitation encountered in the field of MP enumeration. This
was recognized by N. Fisher and colleagues (Abstract 181/B60;
Accurate Detection of Counting Beads for Cell-Derived
Microparticle Enumeration by Flow Cytometry, CYTO 2013).
We did encounter the same problem on the Gallios when
switching from a 0.5 mm eq. down-to a 0.3 mm eq. FSC thresh-
old. In the new MP-optimized FSC settings, the previously
used 10 mm counting beads BC Flow-CountTM were not prop-
erly detected. One possible explanation is that under MP set-
tings, such big beads provide too high scatter signals for the
electronics to detect both the arrival and the disappearance of
each signal. This problem was solved on the Gallios using
smaller beads i.e., 5 mm DAKO CytoCountTM. The same type
of limitation may have appeared during the present study on
the Fortessa using these 5 mm beads with our low SSC thresh-
old, suggesting that smaller counting beads are preferable.
This limitation was circumvented by the use of the “bright
156
PMP” subset as an internal reference in shared samples.
Indeed, most of the bright PMP are also large PMP falling
over 0.5 mm beads in FSC and thus their number should not
be influenced by variations of threshold at the level of 0.3
mm eq. This was confirmed by the correlation observed
between instruments when counting “large & bright PMP”
(data not shown).
Using the same Fortessa with both FSC and SSC capabil-
ities and a low SSC threshold, we achieved conditions permit-
ting acquisition of more PMP than usually feasible. Later we
could complete the selection using scatter-based gates. Since
one of the Fortessa-based platforms had the required mini-
mum resolution to apply the same FSC-based standardization
strategy than on BC Gallios, it was possible to apply our usual
0.3 mm eq. MP gate. Comparing two platforms on the same
instrument did not require internal controls for counting and
it permitted us to find an optimal location for the previously
unknown SSC equivalent of the FSC 0.3 mm eq. threshold.
This is close to 0.17 mm and can be reproducibly located using
a simple formula based on the relative distance between 0.16
and 0.20 mm reference beads. Later, the same FSC-based strat-
egy applied on two different instruments provided very simi-
lar PMP counts. Finally, the major objective of the study was
attained when we demonstrated equivalent PMP counts with
two instruments using two different size-related parameters.
Our study demonstrates that the sizes of reference beads
are clearly different in each strategy. This is depicted in Figure
6, which provides the authors’ vision of the problem. To sum-
marize: (i) in FSC the threshold can be set at 0.3 mm eq and
the MP gate goes from 0.3 mm eq up to 1 mm eq., with a
dichotomy between small and large PMP around 0.5 mm-eq.;
(ii) in SSC, the threshold can be set at about 0.17 mm eq and
the MP gate ranges from 0.17 up to 0.6 mm eq., with a
dichotomy between small and large PMP around 0.22 mm eq.
Thus, although SSC seems to more easily resolve very small
beads between 100 and 200 nm in a group of FCs, it is not
intrinsically a more sensitive size-related parameter than FSC
used in the other group of FCs. Both parameters need to be
compared on the same biological objects and not only on
beads.
Megamix-Plus beads are uniquely appropriate to fit with
the proposed standardization approach since (i) their sizes
have been chosen for a specific purpose, (ii) their fluorescence
intensities were selected to fit with MP-optimized settings,
(iii) they are prepared as ready-to-use cocktails, (iv) the beads
numerical ratios help to setup threshold for some instruments
with discontinuous threshold values (e.g., BC Gallios/Navios),
(v) they display special features helping to locate bead subsets
and avoiding interferences between singlets and doublets, (vi)
they can be easily mixed to create a richer cocktail called Giga-
mix for FCs of high resolution. Although many different sour-
ces of fluorescent size-calibrated beads are available,
comparisons with alternative existing systems demonstrate
their unique adaptation to this specific field of research, as
illustrated in Supporting Information Figures 7 and 8.
The present study suggests a general approach for
quality-control of FCs that is to be used for standardized MP
Standardization of Microparticle Counts

counting. This is summarized in the decision-tree proposed in
Supporting Information Fig. 9. A first step of fluorescence
triggering allows measuring scatter resolutions with FSC- or
SSC-oriented beads (i) to select the most appropriate scatter
parameter in a new unknown instrument or (ii) to regularly
check performance for an already oriented FC. Next, a trigger
is made using the optimal scatter at the proposed standard
threshold level and background noise can be evaluated and
compared to the data treatment capacity of the instrument’s
electronics. For example, a high-end FC such as BD Fortessa
with a theoretical capacity to analyze 40,000 events/s can
more easily deal with a nonspecific background of e.g., 3,000
events/s than a standard FACS-Canto II with a theoretical
electronic capacity to treat 10,000 events/s. Coincidence rates
and thus the risk of swarming effect are different in both cases.
Supporting Information Table 2 provides a list of tentative
speed limits that we consider should not be exceeded when
reproducible MP counts are the aim. Obviously, such a test
operated with filtered water will result into variable back-
ground noise (events/s) between instruments even if they are
the same brand and model. The cleanliness of sheath fluid is
just one factor, the relative intensity of diffracted laser light
may be another variable linked to the optical status.
This study also had some limitations. First, it has been
restricted to the comparison of MP counts on PMP, the most
frequent and well discriminated MP subset found in plasma
samples. This should be extended to more difficult/less repre-
sented MP subsets including erythrocyte-derived and
leukocyte-derived MP. Second, direct generalization regarding
most FCs needs further validation. Third, our standardization
effort has been focused on synchronizing the scatter-based
MP gates. Harmonization of fluorescence detection conditions
requires more efforts than considered in this study, including
Q & B approaches. Fourth, the use of polystyrene beads as size
calibration systems is disputed due to potential differences in
refractive index between beads and cell-derived MP
(18,20–23). Although in our experience using giant viruses of
200 and 400 nm (21), the FSC-based size calibration of these
biological entities was close to values derived from recognized
alternatives such as electron microscopy, the present study
reinforces the notion that plastic beads cannot be used to reli-
ably measure absolute size of MP. We now know that PMP
will receive very different size values whether they are meas-
ured using SSC- or FSC-calibrated scales. Fifth, the use of
scatter for triggering MP analysis is also disputed and several
groups advocate for the use of fluorescence as an optimal trig-
ger (24). Although the sensitivity of FCM is better for fluores-
cence than scatter signals, triggering on fluorescence
encounters several limitations. There are no generic clear-cut
marker available to be used as a biological stain to trigger on
fluorescence. Even Annexin V, one of the highest affinity
potential generic marker for MP is not binding to all MP
(16,25,26). Even though, it would not provide clear-cut
enough staining of the smallest MP to permit discrimination
from background with a no-wash protocol, which is manda-
tory for counting MP in plasma samples. Ultimately, the two
approaches are not mutually exclusive and in the future,
Cytometry Part A  89A: 148 158, 2016
Original Article
maybe there will be options for standardized scatter-based MP
gates to be applied after fluorescence triggering.
To our knowledge, this study is the first to provide a
solution to MP count standardization at this level of sensitiv-
ity. Most of its features should be seriously considered for any
future multi-centric inter-lab comparison, except two of the
six specific points cited above. First the simultaneous analysis
of the same stained samples on two FCs was typical of a
mono-centric exercise. Second the use of large PMP as inter-
nal standard should not be needed provided that smaller
counting beads are selected. The here-proposed strategy is
currently under evaluation in an ISTH-sponsored multicentre
study involving different instruments (both FSC- and/or SSC-
optimized) with common reagents and protocols to count
PMP.
AKNOWLEDGMENTS
The authors wish to thank C. Maier (BD France, Greno-
ble, F) for the generous loan of a fully equipped LSR-Fortessa
instrument for two weeks as well as the expert support of C.
Gameiro (ex BD training center, Rungis, France). External
access to other instruments for preparation of the present
study has been facilitated mainly (a nonexhaustive list) by F.
Mullier (Mt-Godine, B), A.M. Faussat and E.C. Guerin (both
in Paris, F), D. Raoult (Marseille, F), A. Spittler (Vienna, A),
N. Fisher, M. Mooberry and N. Key (Chapel-Hill, NC) as well
as R. Zucker (Research Triangle Park, NC). Noteworthy, apart
from the Gallios and Fortessa FCs, which are central to the
present study, other instruments were only available for a
short time, sometimes only in the course of a demo, and may
not have been tested in their optimal status.
AUTHORSHIP AND DISCLOSURES
PP designed the study, performed the research, analyzed
the data and wrote the manuscript. SR performed the
research, analyzed the data and made the figures. JB and TB
contributed to the flow cytometry experiments. RL supervised
the research and revised the manuscript. FDG participated to
the revision of the manuscript.
PP, TB, JB are full-time employees of the company Bio-
Cytex who developed and sells the Megamix line of products.
Other authors declare no competing financial interests.
LITERATURE CITED
1. Piccin A, Murphy WG, Smith OP. Circulating microparticles: Pathophysiology and
clinical implications. Blood Rev 2007;21:157–171.
2. Burnier L, Fontana P, Kwak BR, Angelillo-Scherrer A. Cell-derived microparticles in
haemostasis and vascular medicine. Thromb Haemost 2009;101:439–451.
3. Wu ZH, Ji CL, Li H, Qiu GX, Gao CJ, Weng XS. Membrane microparticles and dis-
eases. Eur Rev Med Pharmacol Sci 2013;17:2420–2427.
4. Lacroix R, Robert S, Poncelet P, Dignat-George F. Overcoming limitations of micro-
particle measurement by flow cytometry. Semin Thromb Hemost 2010;36:807–818.
5. Lacroix R, Robert S, Sabatier F, Dignat-George F. Flow cytometry. In: Harrison P,
Gardiner C, Sargent I, editors. Extracellular Vesicles in Health and Disease. Singa-
pore: Pan Stanford Publishing; 2014. pp 201–222.
6. Zhu S, Ma L, Wang S, Chen C, Zhang W, Yang L, Hang W, Nolan JP, Wu L, Yan X.
Light-scattering detection below the level of single fluorescent molecules for
high-resolution characterization of functional nanoparticles. ACS Nano 2014;8:
10998–1006.
7. Sarlon-Bartoli G, Bennis Y, Lacroix R, Piercecchi-Marti MD, Bartoli MA, Arnaud L,
Mancini J, Boudes A, Sarlon E, Thevenin B, et al. Plasmatic level of leukocyte-
derived microparticles is associated with unstable plaque in asymptomatic patients
with high-grade carotid stenosis. J Am Coll Cardiol 2013; 62:1436–41.
157
Original Article
8. Nozaki T, Sugiyama S, Koga H, Sugamura K, Ohba K, Matsuzawa Y, Sumida H,
Matsui K, Jinnouchi H, Ogawa H. Significance of a multiple biomarkers strategy
including endothelial dysfunction to improve risk stratification for cardiovascular
events in patients at high risk for coronary heart disease. J Am Coll Cardiol 2009;54:
601–608.
9. Amabile N, Guerin AP, Tedgui A, Boulanger CM, London GM. Predictive value of
circulating endothelial microparticles for cardiovascular mortality in end-stage renal
failure: a pilot study. Nephrol Dial Transplant 2012;27:1873–1880.
10. Robert S, Poncelet P, Lacroix R, Arnaud L, Giraudo L, Hauchard A, Sampol J,
Dignat-George F. Standardization of platelet-derived microparticle counting using
calibrated beads and a cytomics FC500 routine flow cytometer: A first step toward
multicenter studies? J Thromb Haemost 2009;7:190–197.
11. Lacroix R, Robert S, Poncelet P, Kasthuri RS, Key NS, Dignat-George F. Standardiza-
tion of platelet-derived microparticle enumeration by flow cytometry with calibrated
beads: Results of the international society on thrombosis and haemostasis SSC col-
laborative workshop. J Thromb Haemost 2010;8:2571–2574.
12. Nielsen MH, Beck-Nielsen H, Andersen MN, Handberg A. A flow cytometric
method for characterization of circulating cell-derived microparticles in plasma.
J Extracell Vesicles 2014;3:
13. Nebe-von-Caron G. Standardization in microbial cytometry. Cytometry A 2009;75A:
86–89.
14. Macey MG, Enniks N, Bevan S. Flow cytometric analysis of microparticle phenotype
and their role in thrombin generation. Cytometry B: Clin Cytom 2011;80:57–63.
15. Robert S, Lacroix R, Poncelet P, Harhouri K, Bouriche T, Judicone C, Wischhusen J,
Arnaud L, Dignat-George F. High-sensitivity flow cytometry provides access to
standardized measurement of small-size microparticles–brief report. Arterioscler
Thromb Vasc Biol 2012;32:1054–1058.
16. Rousseau M, Belleannee C, Duchez AC, Cloutier N, Levesque T, Jacques F, Perron J,
Nigrovic PA, Dieude M, Hebert MJ, et al. Detection and quantification of micropar-
ticles from different cellular lineages using flow cytometry. Evaluation of the impact
of secreted phospholipase a2 on microparticle assessment. PLoS One 2015;10:
e0116812.
158
17. Lacroix R, Judicone C, Poncelet P, Robert S, Arnaud L, Sampol J, Dignat-George F.
Impact of pre-analytical parameters on the measurement of circulating micropar-
ticles: Toward standardization of protocol. J Thromb Haemost 2012;10:437–446.
18. van der Pol E, van Gemert MJ, Sturk A, Nieuwland R, van Leeuwen TG. Single vs.
Swarm detection of microparticles and exosomes by flow cytometry. J Thromb
Haemost 2012;10:919–930.
19. Nolan JP, Stoner SA. A trigger channel threshold artifact in nanoparticle analysis.
Cytometry A 2013;83A:301–305.
20. Mullier F, Bailly N, Chatelain C, Dogne JM, Chatelain B. More on: Calibration for
the measurement of microparticles: Needs, interests, and limitations of calibrated
polystyrene beads for flow cytometry-based quantification of biological micropar-
ticles. J Thromb Haemost 2011;9:1679–1681; author reply 1681.
21. Robert S, Poncelet P, Lacroix R, Raoult D, Dignat-George F. More on: Calibration
for the measurement of microparticles: Value of calibrated polystyrene beads for
flow cytometry-based sizing of biological microparticles. J Thromb Haemost 2011;9:
1676–1678; author reply 1681–1682.
22. Chandler WL, Yeung W, Tait JF. A new microparticle size calibration standard for
use in measuring smaller microparticles using a new flow cytometer. J Thromb Hae-
most 2011;9:1216–1224.
23. Parida BK, Garrastazu H, Aden JK, Cap AP, McFaul SJ. Silica microspheres are supe-
rior to polystyrene for microvesicle analysis by flow cytometry. Thromb Res 2015;
S0049-3848(15)00066.
24. Arraud N, Gounou C, Linares R, Brisson AR. A simple flow cytometry method
improves the detection of phosphatidylserine-exposing extracellular vesicles.
J Thromb Haemost 2015;13:237–247.
25. Connor DE, Exner T, Ma DD, Joseph JE. The majority of circulating platelet-derived
microparticles fail to bind annexin V, lack phospholipid-dependent procoagulant
activity and demonstrate greater expression of glycoprotein ib. Thromb Haemost
2010;103:1044–1052.
26. Perez-Pujol S, Marker PH, Key NS. Platelet microparticles are heterogeneous and
highly dependent on the activation mechanism: Studies using a new digital flow
cytometer. Cytometry A 2007;71A:38–45.
Standardization of Microparticle Counts