Magnesium Sulfate As An Alternative in Vitro
Magnesium Sulfate As An Alternative in Vitro
Magnesium Sulfate As An Alternative in Vitro
From the Rostock University Medical Centre, Institute of Clinical Chemistry and Laboratory Medicine, Rostock, Germany.
Key Words: Platelet count; Mean platelet volume; Preanalytics; Cell swelling; Anticoagulation; Magnesium sulfate; EDTA; Citrate
DOI: 10.1093/AJCP/AQW066
806 Am J Clin Pathol 2016;145:806-814 © American Society for Clinical Pathology, 2016. All rights reserved.
DOI: 10.1093/ajcp/aqw066 For permissions, please e-mail: [email protected]
AJCP / ORIGINAL ARTICLE
experiences with MgSO4 in patients with PTCP led to the 9NC; fill volume of blood 2.9 mL), and MgSO4 (S-Monovette
commercial launch of a blood sampling tube prepared with ThromboExact; fill volume of blood 2.7 mL), all manufac-
MgSO4 as an anticoagulant. In a multicenter study10 of anti- tured by Sarstedt.
coagulant-induced PTCP, magnesium anticoagulation was The order of blood sampling was the following: (1)
shown to be effective, although platelet counts and platelet EDTA, (2) citrate, and (3) MgSO4. Blood from the healthy
volume seemed to be underestimated compared to when volunteers was used for the above-described time kinetics
using EDTA blood.11 and measured within 5 to 10 minutes after sampling.
The real platelet volume is still a matter of controversy, Patients’ blood sampling was performed in the casualty
especially if one takes into account that EDTA as well as units of the hospital, located either in the same or in the
citrate anticoagulation leads to early platelet swelling.5 This nearby buildings. Due to this fact, the time period between
is a critical issue of preanalytical standardization for meas- sampling and measurement was between 30 and 45 minutes.
uring the mean platelet volume (MPV) as a possible but
controversially discussed predictive biomarker.12-14
mean was used for plotting the results of comparing the im- 400
pedance and the fluorescent optical method for platelet
350
count in differently anticoagulated blood samples.
250
Results
200
Time Dependency of the Measurement of Platelet
150
Parameters in EDTA-, Citrate-, and MgSO4-
Anticoagulated Blood Samples Within 180 Minutes of 100
Venipuncture
Platelet count and MPV of 21 volunteers were meas- 50
14
A
13
250
Platelet Count (× 109/L)
240
12
230
220
MPV (fL)
11
210
200
190 10
180
170 9
160
0 20 40 60 180 8
Time After Sampling (min)
7
20 500
18
400
300
PDW (fL)
14
12 200
10
100
8
6 0
EDTA Citrate MgSO4
Figure 4 Platelet distribution width (PDW) measured by the Figure 5 Comparison of platelet counts in 834 blood sam-
XE 5000 (impedance method) in samples of 21 volunteers ples drawn at the same time but using different anticoagu-
anticoagulated with EDTA (black), citrate (light gray), and lants and measured by the impedance and the optical
magnesium sulfate (dark gray). methods using the XE 5000 as a routine analyzer. The results
of the platelet counts are depicted by box plots as described
under statistical methods (PLTimp, blank; PLTopt, gray).
MgSO4, magnesium sulfate.
(PLTimp) and fluorescence optical (PLTopt) method of the
XE 5000 Figure 5 and Table 1 in parallel. The mean platelet counts measured in citrate and EDTA showed a
platelet count in EDTA-anticoagulated blood was 243.6 slightly worse correlation (r ¼ 0.932; y ¼ 0.83x þ 0.15).
109/L (32-769 109/L) using the impedance method. The
mean platelet count measured by the optical method of the
XE 5000 was 236.5 109/L (39-799 109/L).
The mean PLTimp count for citrate-anticoagulated MPV Measured in Blood Samples Anticoagulated by
blood was 171.2 109/L (24.2-585.2 109/L), although EDTA, Citrate, and MgSO4
the dilution effect had already been corrected by a factor of Platelets anticoagulated with EDTA, citrate, or MgSO4
1.1. The mean value for PLTopt was 195.7 109/L (30.8- differ markedly by volume when measured with the Sysmex
756.8 109/L). XE 5000. The values are depicted as box plots in Figure 6 .
In blood samples anticoagulated with magnesium salt, The respective medians and 5th, 25th, 75th, and 95th per-
the mean PLTimp was 205 109/L (32-642 109/L), and centiles of the platelet counts are summarized in Table 1.
the respective mean value for PLTopt was 226.5 109/L The MPV in EDTA-anticoagulated samples from 834
(34-787 109/L). The differences between PLTimp and patients was 10.4 fL with a minimum of 7.9 fL and a max-
PLTopt were statistically significant independent from the imum of 14.0 fL. In samples from the same patients
anticoagulant used (EDTA, P < .001; citrate, P < .001; anticoagulated with citrate, the mean volume amounted to
MgSO4, P < .001). These findings are depicted as box plots 9.9 fL, ranging from 7.8 to 13.0 fL, and the mean volume in
in Figure 5. magnesium-anticoagulated samples was 9.4 fL with a range
The platelet counts from EDTA-anticoagulated samples from 7.3 to 12.4 fL. These differences in MPV measured in
and those measured from parallel samples anticoagulated by differently anticoagulated whole blood are all highly signifi-
citrate or MgSO4 were correlated by Spearman rank-order cant (P < .001).
correlation. The MPVs measured in MgSO4- and citrate-
Platelet counts measured by impedance in EDTA blood anticoagulated blood were correlated with the respective
(PLTimp) correlated better with platelet counts measured in counts in EDTA as a reference. The correlation coefficient r
MgSO4-anticoagulated blood than platelet counts from cit- for MPVs measured in MgSO4- or EDTA-anticoagulated
rate blood, r ¼ 0.967 (y ¼ 0.80x þ 9.48) and r ¼ 0.834 blood was better than the correlation of the MPVs from
(y ¼ 0.64x þ 15.28), respectively. citrated blood and from EDTA blood (r ¼ 0.902 and
Using the fluorescence optical method, the platelet r ¼ 0.859, respectively). The corresponding regression lines
counts (PLTopt) in MgSO4- and EDTA-anticoagulated blood are defined by the equations y ¼ 0.78x þ 1.24 and
were correlated with r ¼ 0.977 (y ¼ 0.95x þ 2.46), whereas y ¼ 0.78x þ 1.74.
Table 1
Summarized Results of Platelet Counts Measured by Impedance (PLTimp) or Fluorescent Optical Technique (PLTopt) and Mean
Platelet Volume (MPV) From Whole-Blood Samples Anticoagulated by EDTA, Citrate, or Magnesium Sulfate (MgSO4) From 834
Individuals
EDTA Citrate MgSO4
PLTimp, PLTopt, PLTimp, PLTopt, PLTimp, PLTopt,
Characteristic 109/L 109/L MPV, fL 109/L 109/L MPV, fL 109/L 109/L MPV, fL
Minimum 32.0 39.0 7.9 24.2 30.8 7.8 32.0 34.0 7.3
5% 139.7 134.7 9.1 86.5 102.3 8.7 115.7 129.3 8.3
25% 193.0 186.0 9.8 132.3 150.7 9.3 164.0 178.0 8.8
Median 229.0 223.5 10.4 165.0 187.0 9.9 194.0 214.0 9.3
75% 284.0 276.0 11.0 201.3 224.4 10.4 238.8 263.0 9.8
95% 380.7 367.4 12.2 267.7 310.2 11.3 319.4 353.1 10.9
Maximum 769.0 799.0 14.0 585.2 756.8 13.0 642.0 787.0 12.4
14 100
80
12
CD62+ Platelets (%)
60
MPV (fL)
10
40
8
20
6 6
EDTA Citrate MgSO4 7-9.9 10-10.9 ≥11
MPV (fL)
Figure 6 Mean platelet volume (MPV) measured by the im-
Figure 7 Expression of platelet activation markers deter-
pedance method (Sysmex XE 5000) in samples of 834 pa-
mined in whole blood from 21 volunteers and anticoagulated
tients anticoagulated with EDTA, citrate, or magnesium
with different anticoagulants (EDTA, triangle; magnesium
sulfate (MgSO4). The averaged results are depicted by box
sulfate, square; citrate, dot). MPV, mean platelet volume.
plots as described under statistical methods.
Discussion
MPV and Spontaneous Platelet Activation Today, routine platelet counts and volume measure-
It has been reported that the MPV correlates with plate- ments are carried out from peripheral venous blood,
let activation. We therefore measured the spontaneous ex- anticoagulated with EDTA or citrate, using an automated
pression of activation markers on platelets anticoagulated hematology analyzer. An important requirement for the cor-
with all three anticoagulants. In EDTA-anticoagulated rect determination of platelet count and volume is the stand-
blood, the platelets express CD62 to a variable degree, ardization of both, the preanalytical and the analytical
whereas in MgSO4- and citrate-anticoagulated blood, only procedure.17 There is a large body of evidence that the
slight activation is measurable. In Figure 7 , the measured blood-collecting conditions, including the choice of anti-
activity is plotted for three different MPV tertiles (7-9.9 fL/ coagulant, are critical for the reliability of platelet count,
10-10.9 fL/11 fL). It is obvious that the spontaneous activ- MPV,18 and platelet activation.19
ity of EDTA-anticoagulated platelets shows a weak but not On the basis of our experiences with MgSO4 in individ-
significant correlation with the MPV, whereas the otherwise uals with anticoagulant-induced in vitro platelet aggrega-
anticoagulated platelets do not. tion,10 so-called pseudothrombocytopenia, we studied the
suitability of a recently introduced blood-sampling device, This might be explained by a loss of platelets when they
prefilled with MgSO4 (ThromboExact-Monovette; Sarstedt) are in contact with citrate, although there were no signs of
for the measurement of platelet count and volume in com- cell destruction. Another explanation might be the fact that
parison to the established anticoagulants EDTA and citrate. the MPV is markedly lower in citrate-anticoagulated blood,
The rationale and aim of the study was to evaluate the pos- so very small platelets may bypass the lower platelet dis-
sible advantages of MgSO4 as an in vitro anticoagulant, es- criminator. However, this does not explain the low platelet
pecially under the aspect of some shortcomings of the counts measured by the fluorescence method.
established anticoagulants. EDTA is an irreversible calcium chelator, whereas cit-
The cell-swelling and shape-changing effect of EDTA rate binds calcium in a pH-dependent manner. The release
within the first hours after venipuncture is known from the of calcium might lead to the formation of platelet aggregates
literature20,21 and represents one of the major problems in during the storage and repeated agitation of citrate-
standardizing platelet volume measurement. anticoagulated blood. Aggregated platelets are then counted
Lower platelet counts in magnesium-anticoagulated There are some indications from other investigators
blood using the routine impedance method underestimate that the MPV is related to the degree of platelet activation.35
the EDTA platelet count by 5% to 10%. At first glance, this This is in accordance with our findings that the spontaneous
might be critical for patients with thrombocytopenia at a expression of CD62p and CD63 in EDTA-anticoagulated
defined critical transfusion threshold of 109/L, for ex- blood tends to be correlated with the MPV. In contrast, cit-
ample.31 This low threshold was set with the intention to re- rate and especially MgSO4 anticoagulation is accompanied
duce the need for and frequency of platelet transfusion only by marginal spontaneous activation; nevertheless, the
because of economic reasons. In such a situation, platelets platelets can still be activated by suitable agonists (data not
from magnesium-anticoagulated blood have to be measured shown).
by the fluorescent optical method. At least, underestimation We think that the reliability of the MPV is such an im-
of platelet counts is not disadvantageous for the individual portant prerequisite for its use as a predictive and prognostic
patient. biomarker that MgSO4 might be advantageous for this issue;
5. Lancé MD, van Oerle R, Henskens YMC, et al. Do we need 22. Ahnadi CE, Chapman ES, Lépine M, et al. Assessment of
time adjusted mean platelet volume measurements? Lab platelet activation in several different anticoagulants by the
Hematol. 2010;16:28-31. Advia 120 haematology system, fluorescence flow cytometry,
6. Perrotta G, Roberts L, Glazier J, Schumacher HR. Use of and electron microscopy. Thromb Haemost. 2003;90:940-948.
sodium citrate anticoagulant for routine haematology ana- 23. Heilmann EJ, Kundu SK, Sio R, et al. Brief communication
lysis on the CELL-DYNV R 4000: an opportunity to enhance comparison of four commercial citrate blood collection sys-
efficiency in the clinical laboratory. Lab Hematol. tems for platelet function analysis by the PFA-100TM system.
1998;156-162. Thromb Res. 1997;87:159-164.
7. Sahli H. Lehrbuch der klinischen Untersuchungs-Methoden für 24. Narasimha A, Kumar H, Prasad CS. Anticoagulant induced
€
Studierende und praktische Arzte mit 291 teilweise farbigen artefacts in peripheral blood smears. Indian J Hematol Blood
Holzschnitten im Texte und 5 lithographierten Tafeln/von H. Transfus. 2008;24:43-48.
Sahli. 4th ed. Wien, Leipzig: Deutike; 1905. 25. Shechter M, Merz CN, Paul-Labrador M, et al. Oral magne-
8. Nakamoto K, Sugibayashi S, Takahashi A, et al. Platelet sium supplementation inhibits platelet-dependent throm-
count in EDTA-dependent pseudothrombocytopenia—ap- bosis in patients with coronary artery disease. Am J Cardiol.
plication of MgSO4 as an anticoagulant. Rinsho Byori. 1999;84:152-156.