British Journal of Haematology, 1999, 106, 801±808
Lipid composition of seven APTT reagents
in relation to heparin sensitivity
S. K I TC HE N , 1 I. C A RTW RI GH T, 2 T. A. L. W O O D S , 3 I. J E NN I NG S 3 A ND F. E. P R ES TO N 1
Shef®eld Thrombosis and Haemostasis Centre, Royal Hallamshire Hospital, Shef®eld,
2
Department of Molecular Biology and Biotechnology, University of Shef®eld,
and 3U.K. NEQAS (Blood Coagulation), Shef®eld
1
Received 4 February 1999; accepted for publication 1 June 1999
Summary. The phospholipid content of different activated
partial thromboplastin time (APTT) reagents was determined and compared to heparin sensitivity. The seven
reagents included were those most widely used amongst
participants of the U.K. National External Quality Assessment Scheme (NEQAS) at the time of study. Heparin
sensitivity was assessed using the APTT ratios obtained by
more than 300 NEQAS participants on ®ve plasmas prepared
from patients receiving unfractionated heparin. The concentrations of three neutral lipids and six phospholipids
present in the seven APTT reagents were determined by
high-performance thin-layer chromatography (HPTLC) and
densitometry. Both the concentrations and the relative
percentages of individual phospholipid components varied
markedly between reagents. The total phospholipid concentration included a 12-fold range from 16 to 205 mg/ml.
Phosphatidylserine (PS) was completely lacking from one
reagent prepared from vegetable material and ranged from 3
to 22 mg/ml in the other six reagents containing extracts
from animal tissue. The concentration of phosphatidylcholine ranged from 3 to 109 mg/ml. There was no demonstrable
relationship between the concentration of any individual
lipid components and heparin sensitivity. However, the
relative percentage phospholipid composition was important
since a lower % of PS or phosphatidylinositol (PI) correlated
with increasing heparin sensitivity.
The most widely used test for heparin monitoring world-wide
is the activated partial thromboplastin time (APTT). In the
U.K. at least 20 different reagents are employed for heparin
dosage monitoring. A number of studies have identi®ed
differences in the responsiveness of different APTT reagents
to heparin (Banez et al, 1980; Shojania et al, 1988; Kitchen
et al, 1996; Kitchen & Preston, 1996).
There are a number of methodological differences which
might contribute to differences between reagents in respect
of heparin responsiveness, including the use of different
activators and activation times. However, there is evidence of
marked variation of the lipid composition of some APTT
reagents which were employed in the 1980s (Stevenson et al,
1986). This variation has important implications for
reagent sensitivity to heparin since it has been shown that
phospholipid concentration is important in some reagents,
decreasing concentration leading to increasing heparin
sensitivity (Barrowcliffe & Gray, 1981). Some phospholipids
are of particular importance in respect of APTT determinations
and it has been shown that phosphatidylserine (PS) has an
important role (Stevenson & Poller, 1982; Comfurius et al,
1994; Zwaal & Schroit, 1997) and that it inhibits or promotes
coagulation depending on its concentration (Slater et al,
1980). Other major lipid components of platelet membranes are phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylinositol (PI) and sphingomyelin (SM)
(Crawford & Scrutton, 1994). To our knowledge, there are no
recent studies of lipid composition of APTT reagents as
presently employed for monitoring of unfractionated heparin
(UFH) therapy.
The aim of the present study was to characterize the
lipid composition of a number of APTT reagents in routine
use for monitoring UFH therapy and to search for relationships between phospholipid composition and heparin
sensitivity.
Correspondence: Dr S. Kitchen, Shef®eld Thrombosis and Haemostasis Centre, Royal Hallamshire Hospital, Glossop Road, Shef®eld
S10 2JF. e-mail: steve.kitchen@csuh.trent.nhs.uk.
Activated partial thromboplastin time reagents. Seven types of
APTT reagent were included in the study. The characteristics
of these reagents are shown in Table I.
q 1999 Blackwell Science Ltd
Keywords: APTT, heparin sensitivity, phospholipids.
MATERIALS AND METHODS
801
S. Kitchen et al
Table I. APTT reagent source and composition.
Reagent
Manufacturer
Abbreviation
Activator
Phospholipid source
Actin FS
Boehringer PTT
Diagen Bell & Alton
Diagen KPS
IL Lyophilized Silica
Manchester
Platelin LS
Dade
Boehringer Mannheim
Diagnostic Reagents
Diagnostic Reagents
Instrumentation Lab
Withington Hospital
Organon Teknika
AFS
B
DB
DK
IL
M
OP
Ellagic acid
Kaolin
Kaolin
Kaolin
Micronized silica
Kaolin
Micronized silica
Soya extract
Animal tissue extract
Animal brain extract
Animal brain extract
Bovine brain extract
Animal brain extract
Animal tissue extract
Lipid analysis was performed on a single lot/batch of ®ve of
the reagents and three different lots/batches of Instrumentation Laboratory micronized silica reagent and Diagen Kaolin
Platelet Substitute reagents.
Lipid analysis of APTT reagents. Each APTT reagent was
reconstituted in distilled water according to the manufacturer's instructions where necessary (reagents AFS and OP
were supplied by the manufacturers in liquid form).
Extraction of lipids. Lipid extraction was performed as
previously described by Cartwright et al (1997). Each APTT
reagent (1 ml of aqueous suspension) was mixed with 20 ml
of chloroform/methanol (2:1 v/v) to give a single phase. This
monophasic system was maintained during centrifugation at
1000 g for 10 min. The supernatant (a single phase) was
removed from the pelleted silica/kaolin debris and shaken
vigorously with 4 ml of 50 mM calcium chloride for 2 min.
This mixture formed a biphasic system following recentrifugation at 2000 g for 5 min with the lower organic
phases containing the lipids and the upper aqueous phase
containing the soluble non-lipid contaminants. This was
then evaporated to dryness and stored at 208C prior to
analysis by HPTLC. We performed recovery experiments by
absorbing a known amount of phospholipid onto silica in
order to mimic APTT reagents. When aqueous suspensions
of these were extracted, as described above, >99% of
phospholipid mass was recovered in a single extraction step.
Lipid extracts were re-dissolved in chloroform/methanol
(1:1, v/v) and 10 ml aliquots applied to nanosil 20 ´ 10 cm
HPTLC plates using a Camlab Desaga AS30 Autospotter.
Phospholipid standards (0´5±20 mg) or neutral lipid standards (0´5±10 mg were also applied to the ®rst ®ve lanes on
each plate. The following lipid standards (Sigma, Poole) were
used: L-a-phosphatidyl choline-dioleoyl; L-a-phosphatidyl
ethanolamine-dioleoyl; L-a-phosphatidyl inositol (ammonium salt); DL-a-phosphatidyl-L-serine dipalmitoyl (sodium
salt); diphospatidyl inositol (sodium salt); diphosphatidylglycerol (sodium salt); sphingomyelin; cholesterol; cholesterol
oleate; triolein.
Lipids were separated using the following developing
solvent systems: propan-1-ol/propionic acid/chloroform/
water (45/30/30/15 by volume) for phospholipids and
heptane/diethylether/glacial acetic acid (60/40/2 by
volume) for neutral lipids. The plates were air-dried and
dipped in cupric acetate (3%, w/v) dissolved in phosphoric
acid (8%, w/v) for 30 s, air-dried again and heated to 1408C
on a hot plate for 30 min. After cooling, the individual lipid
components were determined using a Camlab Desaga CD60
densitometer in re¯ectance mode at 340 nm. The following
phospholipid components were quanti®ed: phosphatidylcholine (PC); phosphatidylethanolamine (PE); phosphatidylinositol (PI); phosphatidylserine (PS); sphingomyelin (SM);
diphosphatidyl glycerol (DPG). Three neutral lipids were also
quanti®ed: cholesterol (C); cholesterol ester (CE); triglyceride
(TG).
Heparin sensitivity of APTT reagents. Data were extracted
from U.K. NEQAS surveys in relation to the seven APTT
reagents. The median APTT ratio (test/mid-point of local
normal range) of participants using these particular reagents
was obtained on each of ®ve surveys. The number of centres
using each reagent varied from 10 to 170. Each sample was
prepared as a pool of citrated plasmas from patients receiving
unfractionated heparin (Pumphep, Leo Laboratories Ltd),
lyophilized and distributed to U.K. NEQAS participants as
previously described (Kitchen et al, 1996).
Heparin sensitivity of different lots/batches of reagent of the
same type. Blood samples were obtained from 20 normal
subjects and two groups of 20 patients receiving unfractionated heparin (Pumphep, Leo Laboratories Ltd) for venous
thromboembolic disease using an evacuated collection
system (Vacutainer, Becton Dickinson Ltd). In each case
4´5 ml blood was mixed with 0´5 ml 0´105 M buffered
trisodium citrate and centrifuged at 2500 g for 10 min at
room temperature. Three lots/batches of reagent DK and
reagent IL were employed for determination of APTT ratio
(patient over mean normal APTT). A different group of
patients was used with each type of reagent to allow
determinations to be performed within 2 h of sample
collection. All samples were tested in duplicate. Reagent
DK was employed with manual technique and reagent IL in
combination with an automated coagulation laboratory
(ACL) instrument
RESULTS
Results of median APTT ratios obtained by NEQAS
participants are summarized in Table II.
Lipid composition of APTT reagents
An example of HTPLC separation of phospholipids in the
APTT reagents is shown in Fig 1. The lipid concentrations of
q 1999 Blackwell Science Ltd, British Journal of Haematology 106: 801±808
13652141, 1999, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2141.1999.01596.x by CochraneChina, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
802
803
Table II. APTT ratios obtained by U.K. NEQAS participants on ®ve plasmas prepared from
heparinized patients.
AFS
1
2
3
4
5
Mean
n
B
DB
DK
IL
M
OP
1´63
1´67
1´66
2´36
1´91
1´30
1´37
1´31
1´77
1´54
1´55
1´59
1´50
2´21
1´85
1´45
1´53
1´47
1´96
1´64
1´93
2´08
2´09
3´14
2´31
1´67
1´73
1´70
2´20
1´93
1´75
2´36
1´93
3´60
2´59
1´85
10
1´46
21
1´74
23
1´61
60
2´31
170
1´85
23
2´45
25
n ˆ average number of participants using that particular reagent. Each ®gure is the mean
APTT ratio (test APTT/mean normal APTT) obtained by centres using that particular reagent in
the analysis of ®ve different samples (numbered 1±5). Reagent abbreviations are as in Table I.
Table III. Concentrations of phospholipids and neutral lipids during determination of APTT using different reagents.
AFS
B
DB
DK
IL
M
OP
PC
PE
PS
PI
DPG
SM
Total
phospholipid
C
CE
TG
109´3
14´4
5´9
7´5
13´4
2´8
58´6
59´1
40´0
21´7
23´2
28´6
8´8
43´0
nd
22´0
9´6
10´5
12´9
3´1
12´2
37´0
0´4
0´45
0´55
1´49
0´11
nd
nd
nd
0´41
0´49
1´0
nd
nd
nd
2´2
8´6
10´4
12´6
1´3
nd
205´5
79´2
46´6
52´6
70´0
16´1
113´9
0´26
9´1
6´97
7´2
0´81
2´25
14´3
0´86
6´11
2´55
0´10
1´24
0´87
5´43
2´19
nd
nd
nd
nd
nd
nd
All ®gures are in mg/ml. Phospholipid abbreviations are: phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS),
phosphatidylinositol (PI), diphosphatidyl glycerol (DPG), sphingomyelin (SM), cholesterol (C); cholesterol ester (CE); triglyceride (TG). APTT
reagent abbreviations are as in Table I.
nd ˆ not detected.
the APTT reagents are shown in Table III. The data are
presented in this form to take account of differences in the
proportions of APTT reagents in the ®nal reaction mixture in
which clotting time was determined (e.g. reagents IL and DB
were diluted 1 in 3 and 1 in 4 respectively). Results in
relation to PC, PS, PE and PI content are shown in Fig 2.
The proportion of each component phospholipid was
calculated as a % of the total phospholipid. These data are
Table IV. Relative percentage phospholipid composition different
APTT reagents. Phospholipid abbreviations are as in Table III. APTT
reagent abbreviations are as in Table I.
AFS
B
DB
DK
IL
M
OP
PC
PE
PS
PI
DPG
SM
53´2
18´3
12´7
14´4
19´1
17´2
51´5
28´8
50´6
46´4
44´1
40´9
54´7
37´8
0
27´8
20´6
19´9
18´4
19´5
10´7
18´0
0´5
1´0
1´1
2´1
0´7
0
0
0
0´9
0´9
1´4
0
0
0
2´8
18´4
19´6
18´1
7´9
0
shown in Table IV and also in Fig 3 (PC, PS, PE, PI, DPG and
SM). The phospholipid and neutral lipid composition varied
markedly between reagents. For example, the total phospholipid concentration included a 12-fold range (16±
205 mg/ml). Phosphatidylserine was undetectable in one
reagent and ranged from 3 to 22 mg/ml in the other six
reagents. The ®nal concentration of PC in the APTT reaction
mixture included a 35-fold range. Many other differences
between reagents were present.
Batch-to-batch variation in lipid composition of two reagents
Three different batches/lots of two types of APTT reagent
were available for phospholipid analysis. Results of the
analysis are shown in Table V. All concentrations are in mg/
ml ®nal concentration in the APTT reaction mixture. The
proportion of each component phospholipid as a % of the
total phospholipid present is shown in Table VI.
For both reagents there was approximately a 20%
difference between batches in respect of ®nal concentration
of phospholipid. The percentage phospholipid composition of
each batch was similar in the case of both types of reagent.
For DK the mean APTT ratios of 20 samples from
heparinized patients were in the range 2´21±2´24 (analysis
q 1999 Blackwell Science Ltd, British Journal of Haematology 106: 801±808
13652141, 1999, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2141.1999.01596.x by CochraneChina, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Heparin Sensitivity and Lipid Composition of APTT Reagents
S. Kitchen et al
Fig 1. HPTLC separation of phospholipids in APTT
reagents (lanes 6±16). Lanes 1±5 contain
increasing concentrations of standard phospholipids. Phospholipid abbreviations are: phosphatidylcholine (PC); phosphatidylethanolamine
(PE); phosphatidylserine (PS); sphingomyelin (SM).
of variance, not signi®cant) for three different batches/lots of
reagent (Table V). For IL reagent (and a different group of
patients) APTT ratios with one lot were signi®cantly greater
than with each of the other two (ANOVA, P < 0´05). Mean
ratios were 2´83, 2´67 and 2´57 (Table V).
Relationship between lipid composition and heparin sensitivity
The relationship between heparin sensitivity of reagents and
the lipid composition of reagents was evaluated by calculating the correlation coef®cient for the mean APTT ratio as
obtained with each reagent by NEQAS participants (on ®ve
plasmas), and the lipid components of those same reagents.
A signi®cant positive or negative correlation would then
indicate the presence of a relationship between heparin
sensitivity and lipid component.
When ®nal concentrations (in mg/ml) of each lipid
component were compared to mean APTT ratio there was
no signi®cant correlation for any of the phospholipids or
neutral lipids individually, or for the total phospholipid
concentration. The correlation coef®cients (r) were in the
q 1999 Blackwell Science Ltd, British Journal of Haematology 106: 801±808
13652141, 1999, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2141.1999.01596.x by CochraneChina, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
804
805
Fig 2. Phospholipid concentrations of seven different APTT reagents.
Fig 3. Relative % phospholipid composition of seven APTT reagents.
range 0´13 to 0´29, with P values all >0´50. This suggests
a complete absence of any relationship in respect of
concentrations, at least for the range present in these
reagents. When the relative proportions of each phospholipid
in % (of total phospholipid present) were compared to
heparin sensitivity (APTT ratio) there was again no
signi®cant correlation, although higher r values were
obtained in some cases (0´50 in respect of % PC, 0´41 in
respect of % PE, and 0´43 for %PS, all not signi®cant).
The reagent AFS, the only vegetable extract, differed
markedly in composition from the other reagents, prepared
as extracts of animal tissues, most notably in the complete
lack of detectable PS. There are structural similarities
between PS and PI and both of these negatively charged
phospholipids are similar in their location at the inner
surface of resting platelet membranes whereas PC, PE and
SM are main components of the outer lea¯et. For this reason
the possibility that the role of PS in some reagents was
undertaken by PI in the AFS reagent was considered. There
was indeed a signi®cant negative correlation (r ˆ 0´79,
P ˆ 0´034) between APTT ratio and the combined percentages of PS and PI. Thus the lower the % of PS or PI the more
sensitive the reagent was to heparin (indicated by higher
APTT ratios; Fig 4). Furthermore when correlations were
re-calculated after exclusion of AFS data there was a
signi®cant negative correlation between heparin sensitivity
Table V. Concentrations of phospholipids and neutral lipids during determination of APTT for three different batches/lots of two different
reagents. All ®gures are in mg/ml. Phospholipid abbreviations are as in Table III; APTT reagent abbreviations are as in Table I.
DK-1
DK-2
DK-3
IL-1
IL-2
IL-3
PC
PE
PS
PI
DPG
SM
Total
phospholipid
C
CE
TG
APTT
ratio
7´5
6´3
5´1
13´4
11´2
10´6
23´2
20´2
17´5
28´6
24´4
22´6
10´5
9´5
7´9
12´9
12´8
11´3
0´55
0´55
0´44
1´49
1´23
1´1
0´49
0´53
0´7
1´0
0´78
0´75
10´4
9´9
6´5
12´6
13´8
11´9
52´6
47´0
38´1
70´0
64´2
58´3
7´2
6´1
6´25
0´81
0´12
0´08
0´10
0´20
0´12
1´24
1´48
1´38
nd
nd
nd
nd
nd
nd
2´24
2´21
2´21
2´57
2´67
2´83
APTT ratios are the mean of 20 patients receiving heparin, different patients being used for DK and IL reagents.
q 1999 Blackwell Science Ltd, British Journal of Haematology 106: 801±808
13652141, 1999, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2141.1999.01596.x by CochraneChina, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Heparin Sensitivity and Lipid Composition of APTT Reagents
S. Kitchen et al
Fig 4. Relationship between combined % of PS and PI and heparin
sensitivity, as indicated by the APTT ratio. Each APTT result is the
mean value obtained by NEQAS participants using each different
reagent. The regression line is shown.
and % of PS (r ˆ 0´86, P ˆ 0´028; Fig 5). The signi®cant
negative correlation involving % PS and %PI combined
was present and similar whether AFS was included (see
above) or not (r ˆ 0´82, P ˆ 0´045).
The correlation for a comparison of heparin sensitivity
and the ratio of PC to PS or PC to PS ‡ PI was not signi®cant,
although r values were 0´74 (P ˆ 0´09) for PC : PS excluding
AFS data and 0´72 (P ˆ 0´10) or 0´65 (P ˆ 0´12) for PC:
PS ‡ PI including or excluding AFS.
Table VI. Relative percentage phospholipid composition of three
different batches/lots of two different reagents. Phospholipid and
APTT reagent abbreviations are as in Table III and Table I.
DK-1
DK-2
DK-3
IL-1
IL-2
IL-3
PC
PE
PS
PI
DPG
SM
14´4
13´4
13´4
19´1
17´5
18´2
44´1
43´0
45´9
40´9
38´0
38´8
19´9
20´2
20´7
18´4
19´9
19´4
1´1
1´2
1´1
2´1
1´9
1´9
0´9
1´1
1´8
1´4
1´2
1´3
19´6
21´1
17´1
18´1
21´5
20´4
Fig 5. Relationship between % of PS and heparin sensitivity, as
indicated by the APTT ratio. Each APTT result is the mean value
obtained by NEQAS participants using each different reagent. The
regression line is shown (excluding the reagent AFS, which is shown
as a star).
Exclusion of AFS data did not in¯uence the other
correlations appreciably. Thus there was no signi®cant
correlation between ®nal phospholipid concentrations (in
mg/ml) and heparin sensitivity even in the absence of AFS
data.
DISCUSSION
The purpose of the present study was to determine the
relationship between lipid composition and heparin sensitivity of seven different widely used U.K. APTT reagents. Data
obtained in a large number of different centres who were
participants of the U.K. NEQAS were used as an indication of
heparin sensitivity. The mean APTT ratio for ®ve pooled
plasmas collected from patients receiving heparin was
calculated. The reagents studied were those used by 10 or
more U.K. NEQAS participants and the number of users
ranged from 10 to 170 depending on reagent. Analysis
of lipid constituents of APTT reagents revealed marked
differences between reagents. The total phospholipid concentration included a 12-fold range, whereas the mean
APTT ratio with these reagents varied only between 1´46
q 1999 Blackwell Science Ltd, British Journal of Haematology 106: 801±808
13652141, 1999, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2141.1999.01596.x by CochraneChina, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
806
and 2´45. Individual phospholipid components also varied
markedly with PC concentrations from 2´8 to >100 mg/ml.
Phosphatidylserine was undetectable in Actin FS (AFS), the
only reagent of vegetable origin, and ranged from 3 to 22 mg/
ml in the remaining reagents (extracted from animal
tissues). In a previous study Stevenson et al (1986) reported
a low but measurable concentration of PS in AFS. This
earlier study used a different technique which may explain
the presence of detectable PS, although it is not known
whether the manufacturer has modi®ed the reagent bearing
the same name during the period between the two studies.
None of the other materials evaluated by Stevenson et al
(1996) were widely used amongst U.K. NEQAS participants
during the present study and were therefore not included in
the present study.
In vivo clotting reactions are supported by phospholipids
associated with cell membranes, particularly platelets. It has
been shown that the negatively charged phospholipids PS and
PI are con®ned to the interior of the membrane bilayer, but
after activation PS in particular are translocated to the outer
layer where in conjunction with SM, PE and especially PC they
can support coagulation reactions (Schick et al, 1976; Chap
et al, 1977; Comfurius et al, 1994; Zwaal & Schroit, 1997). It
has further been shown that PS is of particular importance
(Stevenson & Poller, 1982) in this respect.
In order to search for a relationship between phospholipid
components and heparin sensitivity the effect of concentration and also the in¯uence of the relationship between
different lipids were considered. In the present study there
was no demonstrable relationship between the concentration of any of the individual lipid components and the
heparin sensitivity of the reagent, as indicated by APTT
ratios obtained in a series of laboratories. Slater et al (1980)
noted that PS inhibited or promoted coagulation depending
on the concentration, although the main inhibitory effects
were at PS concentrations of 30±1000 mg/ml whereas
APTTs differed by <10% in the PS concentration range
observed in the animal tissue extracts included in the present
study (3±22 mg/ml).
In the present study there was no correlation between the
proportion (%) of the total phospholipid present as PS, PC,
PE, PI or SM and heparin sensitivity. However, the possibility
that correlations were in¯uenced by the inclusion of AFS,
which was fundamentally different from the other reagents,
for example in the complete lack of PS, could not be excluded.
For this reason, and because PS and PI are both negatively
charged and are located in similar areas of the platelet
membrane, the possibility that PI was taking the place of PS
in AFS was evaluated. The combined % of phospholipids
present as either PS or PI was compared to heparin sensitivity.
In this comparison, a statistically signi®cant negative correlation was present, suggesting that a lower percentage of
phospholipid as PS or PI was associated with increasing
sensitivity to heparin (increasing APTT ratio). Furthermore
when AFS was excluded from correlation calculations, a
signi®cant negative correlation between the % of PS and APTT
ratio was present. Again the lower the proportion of PS, the
greater was the sensitivity to heparin. These ®ndings together
suggest that the relative proportion of PS or PS/PI is important
807
rather than the actual concentration, at least over the
concentration range in the reagents studied here. This is
further indicated by the correlation coef®cients of around 0´7
(with P values approaching signi®cance) for the comparison
between the APTT ratio and the ratio of PC to PS and PC to PS/
PI combined.
There is evidence that the type of activator used may
in¯uence heparin sensitivity, at least for some phospholipid
reagents, clotting times being more prolonged when kaolin
was employed compared to ellagic acid (Barrowcliffe & Gray,
1981). In contrast, APTT ratios were more prolonged with
the reagent containing ellagic acid (AFS) than with the four
reagents in which kaolin was activator in the present study.
This suggests that the effect of different activators may
depend on the phospholipid in the reagent. Since the two
most sensitive reagents in our study both employed
micronized silica, it is possible that the activator used also
in¯uenced the observed relationship between heparin
sensitivity.
Variation between batches of the same reagent in respect
of heparin sensitivity have been described (Shojania et al,
1988). Three different production lots of two different types
of APTT reagent were available for study. In both cases the
relative composition of different lots were similar, with
similar % of each phospholipid, although there was a 20%
difference in concentrations in both types of reagent. For one
reagent heparin sensitivity did not exhibit signi®cant lot to
lot variation. For the other reagent studied a small but
statistically signi®cant difference between APTT ratios with
different lots was present. In this case the heparin sensitivity
increased as the total phospholipid concentration decreased.
A similar effect has been described for a different reagent
(Barowcliffe & Gray, 1981).
The present study has con®rmed that the relative
proportion of phospholipid present as phosphatidylserine is
an important determinant of heparin sensitivity of some
APTT reagents. It should be noted, however, that the heparin
sensitivity of a vegetable extracted reagent completely
lacking in PS was similar to that of another reagent in
which PS constituted 20% of the phospholipid. In the
vegetable extract reagent the possibility that PI could
substitute for PS could not be excluded from our data.
Already APTT reagents containing carefully selected phospholipids are becoming available. Where these are employed
for heparin monitoring, PS and PI are likely to be of
particular importance.
ACKNOWLEDGMENT
The assistance of U.K. NEQAS (Blood Coagulation) participants is gratefully acknowledged.
REFERENCES
Banez, E.I., Triplett, D.A. & Koepke, J. (1980) Laboratory monitoring
of heparin therapy: the effect of different salts of heparin on the
APTT. American Journal of Clinical Pathology, 74, 569±574.
Barrowcliffe, T.W. & Gray, E. (1981) Studies of phospholipid reagents
used in coagulation. II. Factors in¯uencing their sensitivity to
heparin. Thrombosis and Haemostasis, 46, 634±637.
q 1999 Blackwell Science Ltd, British Journal of Haematology 106: 801±808
13652141, 1999, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2141.1999.01596.x by CochraneChina, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Heparin Sensitivity and Lipid Composition of APTT Reagents
S. Kitchen et al
Cartwright, I.J., Higgins, J.A., Wilkinson, J., Bellavia, S., Kendrick,
J.S. & Graham, J.M. (1997) Investigation of the role of lipids in the
assembly of very low density lipoproteins in rabbit hepatocytes.
Journal of Lipid Research, 38, 531±545.
Chap, H.J., Zwaal, R.F.A. & Van Deenen, L.L.M. (1977) Action of
highly puri®ed phospholipases on blood platelets: evidence for an
asymmetric distribution of phospholipids in the surface membrane. Biochimica et Biophysica Acta, 467, 146±169.
Crawford, N. & Scrutton, M.C. (1994) Biochemistry of the blood
platelet. In: Haemostasis and Thrombosis, 3rd edn (ed. by A. L.
Bloom, C. D. Forbes, D. P. Thomas and E. G. D. Tuddenham).
Churchill Livngstone, London.
Comfurius, P., Smeets, E.F., Willems, G.M., Bevers, E.M. & Zwaal, R.F.A.
(1994) Assembly of the prothrombinase complex on lipid vesicles
depends on the stereochemical con®guration of the polar headgroup
of phosphatidylserine. Biochemistry, 33, 10319±10324.
Kitchen, S., Jennings, I., Woods, T.A.L. & Preston, F.E. (1996) Wide
variability in the sensitivity of APTT reagents for monitoring
heparin dosage. Journal of Clinical Pathology, 49, 10±14.
Kitchen, S. & Preston, F.E. (1996) The therapeutic range for heparin
therapy: relationship between six APTT reagents and two heparin
assays. Thrombosis and Haemostasis, 75, 734±739.
Schick, P.K., Kurcia, K.B. & Chako, G.K. (1976) Location of
phosphatidylethanolamine and phosphatidylserine in the
human platelet plasma membrane. Journal of Clinical Investigation,
57, 1221±1226.
Shojania, A.M., Tetreault, J. & Turnbull, G. (1988) The variations
between heparin sensitivity of different lots of APTT reagents
produced by the same manufacturer. American Journal of Clinical
Pathology, 89, 19±23.
Slater, P.J., Stevenson, K.J. & Poller, L. (1980) Procoagulant activity
of partial thromboplastin (cephalin) reagent and its phospholipid
composition. Thrombosis Research, 18, 831±838.
Stevenson, K.J., Easton, A.C., Curry, A., Thomson, J. & Poller, L.
(1986) The reliability of APTT methods and the relationship to
lipid composition and ultrastructure. Thrombosis and Haemostosis,
55, 250±258.
Stevenson, K.J. & Poller, L. (1982) The procoagulant activity of
partial thromboplastin extracts: the role of phosphatidylserine.
Thrombosis Research, 26, 341±350.
Zwaal, R.F.A. & Schroit, A.J. (1997) Pathophysiologic implications of
membrane phospholipid asymmetry in blood cells. Blood, 89,
1121±1132.
q 1999 Blackwell Science Ltd, British Journal of Haematology 106: 801±808
13652141, 1999, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2141.1999.01596.x by CochraneChina, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
808