gave a minimum polarization
span of over 100 arbitrary
(millipolarization)
units.
Controls and specimens. A set of low, medium, and high
controls were prepared the same as, but independently of,
the calibrators, in concentrations of 10.0, 30.0, and 50.0 mg/
L. Clinical specimens were those from patients receiving
only pentobarbital.
300
Polarization, arb.
(millipolarization)units
Results and Discussion
250
The mean analytical recovery, which we assessed in the
range of5.0to 60.0mg/L, was 99.6% (range95-104%). The
usefulanalytical range was 1.0 to 80.0mg/L. The overall
day-to-day CVs (n
21) for the low-, medium-, and highconcentration controls were 4.5, 2.8, and 2.7%, respectively.
We have not compared results by this method with those by
gas-liquid chromatography.
It may be desirable for a Thx user to obtain an authorization code from the unassigned assay slot (e.g., assay 34), but
we have not found it necessary. A limitation of this assay is
that no other barbiturate can be present before and during
monitoring.
=
200
0
10
30
20
40
50
60
10
80
Pentobarbitalconcn, mg/L
Fig. 1. Calibration curve for pentobarbital
References
users. “Barcode override” mode was used in starting up a
run.
Calibration
of instrument.
A six-point calibration curve
was prepared every week during this’ study with the use of
pentobarbital calibrators A through F. The net polarization
for the calibrators did not change appreciably from week to
week. Figure 1 illustrates a typical calibration curve, which
1. Miller JD, Becker DP, Ward JD. Significance of intracranial
hypertension in severe head injuries. J Neurosurg 47, 503 (1977).
2. Marshall LF, ShapiroHM, RauscheraA, et al. Pentobarbital
therapy for intracranial hypertension in metabolic coma: Reye’s
syndrome. Crit Care Med 6,1 (1978).
3. Wellington P, Rutkowski RB. Rapid estimation of plasma pentobarbital levels by an enzyme immunoassay for barbiturates (EMrr).
Ther Drug Monit 4, 319 (1982).
4. Anticonvulsant drug assays. TDx Operator’s Manual, Abbott
Laboratories, Irving, TX, 1982.
CLIN. CHEM. 30/2, 308-310 (1984)
Interconversion and Comparison of the Results of Three Methods for
Cholinesterase in Serum
Mary A. Newman and Shane S. Que Kee
We investigated three methods for determination of cholinesterase (ChE) in human blood sera, using reference sera, sera
of known ChE content, and serum variously diluted with
deactivated sera. The methods were the Du Pont aca
analyzer, the Pfizer “ChE-tel” method, and the Michel method with and without direct reagent-blank correction. With
respect to precision and reproducibility the methods ranked:
aca > Michel method corrected directly for reagent pH
changes (Michel) > ChE-tel method > Michel method corrected indirectly for reagent pH changes. The interrelationships derived, in terms of each method’s individual units
were:
Michel
ChE-tel
ChE-tel
=
(0.054 ± 0.001) aca (0.043 ± 0.086), r2 0.947
(4.44 ± 0.30) aca
(5.11± 0.99), r2 0.959
(80.2 ± 5.8) Michel + (0.028 ± 0.240), r2 0.947
-
=
=
-
=
=
=
The ChE-tel method was subject to variations due to the
quality of different kits. The aca results were validated by a
Department of Environmental Health, M.L. #56, University of
Medical Center, 3223 Eden Ave.,Cincinnati,
OH 45267.
Received May 2,1983; accepted November 3, 1983.
Cincinnati
308
CLINICAL CHEMISTRY,Vol. 30, No. 2, 1984
collaborative laboratory study. Sera of known activity can be
used effectively for accurate calibration.
Additional Keyphrases: aca discrete analyzer
reference interval
Eliman reagent
butyryithiocholine
The activities
of cholinesterase
(AChE; EC 3.1.1.7) or
“pseudocholinesterase” (ChE; EC 3.1.1.8) in blood plasma or
serum are measured to gauge exposure to orgqnophosphate
pesticides (1) and to safeguard patients who are to be
exposed to succinylcholine anesthetic (2).
Results of the available methods (3, 4) are usually expressed in units characteristic of each method, and not in
terms of international units. Thus, published values are not
comparable from one laboratory to another, unless interlaboratory studies are also performed simultaneously
with
reference sera. Consequently, the ability of epidemiologists
to assess the significance of the data is also limited.
In this study, we used three methods for ChE or AChE on
the same serum samples. These techniques were the Michel,
the Pfizer ChE-tel, and the Du Pont aca. Equations to
enable conversion between these methods were evolved, and
each method was compared for precision and reproducibility.
Materials and Methods
Venous blood was collected in Vacutainer Tubes from
men and women patients before anesthesia with succinylcholine. After analysis for ChE by the Du Pont aca method
(5), the serum was left at room temperature throughout the
day, then stored refrigerated (4#{176}C)
for no longer than seven
days, when three sets of five blood serum samples in
triplicate were then measured for their AChE or ChE
content within 10 mm by the Michel (6-8), the Pfizer ChEtel (9), and the Du Pont aca methods (5) on the same day.
One set, with ChE values ranging between 2.5 to 20.5 kUIL
(“Actual”), was chosen from refrigerated samples previously
measured in the aca for their ChE content. The second set
(“Deact 1”), containing ChE in the range below detection to
22 kUIL, was produced by diluting a serum with high ChE
activity with high-activity serum that had been fully deactivated by heating at 58 to 62#{176}C
for 1 h. The third set (“Deact
2”), with ChE activity 9 tc. 22 kU/L, was obtained by
diluting a sample of high ChE activity with partly deactivated serum produced by similarly heating the high-activity
serum at 54 to 58#{176}C.
A reference serum (Warner Lambert,
Morris Plains, NJ 07950) was also included with each set, as
well as blanks. The samples were only identified to the
analyst by numbers, and arranged randomly (10).
We performed selected measurements in the middle ChE
range five times, to determine how many replicates were
necessary for precise results.
For the Michel method (7), a 0.1-mL serum sample was
analyzed. Reagent controls utilized 0.1 mL of distilled
water. The results were processed in two ways. One method
(Michel 1) followed that proposed by Michel (6) and by NIOSH
(7), and the other (Michel 2) was the one recommended by
the Federal Aviation Administration
(FAA) (8). For Michel
2, the pH change in the reagent control was accounted for by
direct measurement rather than from theoretical considerations as is the case in Michel 1.
For the Pfizer ChE-tel method (9), we used a kit (Pfizer
Diagnostics, 300 W. 43rd St., New York, NY) and 20-.L
samples of serum. The kit is based on the method of Garry
and Routh (11), which is a modification of the original
method of Ellman et al. (12). The quality of the 5’,5’dithiobis(2-nitrobenzoic
acid) solutions was assessed by the
method of Ellman et al. (12), by using reduced glutathione.
For the Du Pont aca method (5), we utilized Du Pont aca
test packs (Du Pont Instruments-aca
Div., Du Pont Co.,
Wilmington, DE 19898), serum being placed in the analyzer
receptacle of the Model H Automatic Clinical Analyzer. This
method was first published by Gal and Roth (13).
Results
Many hospital laboratories assay just one sample instead
of replicates. In testing the validity of making only one
measurement of ChE by the various techniques, we found
that the initial measurement by the aca, ChE-tel, Michel 1,
and Michel 2 techniques differed from the mean of five
replicates by 1.2 (SD 0.33), 3.33 (SD 1.68), 1.78 (SD 0.53),
and 1.93 (SD 0.72) %, respectively-all
acceptable results.
Triplicate measurements appeared to be as precise as those
of five replicates. The aca method is the most precise,
followed by the Michel methods and then the ChE-tel
technique for a given serum sample.
The relative standard error (RSE), in percent, for the
“between-run” data is lowest for the aca method (0.9%),
followed by Michel 2 (6%), Michel 1 (10%), and then the
ChE-tel (13%), the same order as found for the precision
studies. Two different ChE-tel kits produced statistically
significantly different answers for the same reference serum
(mean ± SE for kit 1: 56.0 ± 3.5; kit 2: 33.0 ± 7.6 ChE-tel
units). Since the qualityof
the Ellman reagent did not vary
significantly,
the poorer precision here was due to kit
variation.
Interlaboratory quality-control data for a reference standard revealed that the present results were acceptable: (10.4
± 0.1 kUIL for the aca method, 0.51 ± 0.03 pH unit/h for
Michel 2,44.5 ± 5.7 ChE-tel units for ChE-tel). The interlaboratory
consensus range for the Michel method (0.3-0.8
pH unit/h) was very broad relative to the aca method (9-12
kU/L), perhaps because different
laboratories
used either
Michel method. The ChE-tel method should show RSEs no
larger than 10 to 15% (9); the present ChE-tel results are
marginally acceptable, and clearly are inferior in comparison with the other methods.
After triplicate measurements by the three techniques of
a high-ChE sample diluted to known amounts by variously
deactivated sera (Deact 1, Deact 2, and Actual), the correlations between nominal values of the aca method (aca) and
those observed (aca) were:
For Deact 1, aca = 1.00 aca,, + 0.41, r2
0.998
For Deact 2, aca
1.01 aca + 0.16, r2
0.998
For Actual, aca
1.03 aca,, - 0.43, r2
0.999
Thus, the correlation between the observed and expected
values is 1:1, and the intercepts are close to zero, although
Deact 1 has a larger intercept value than Deact 2. The Deact
1 and 2 data support the adequacy of dilution with fully
deactivated sera as a quality-control
procedure to ensure
that instrument linearity is maintained. The data for Actual
samples show that the samples were adequately stable over
the seven-day storage period before three-way analysis, a
fact referred to in reference 5.
Figure 1 depicts the evident linear relationships of the
ChE-tel and Michel 2 methods vs the aca method, and’ the
ChE-tel vs the Michel 2 method, respectively, including all
the mean-value data.
Table 1 shows how linearity parameters varied from
method to method. Since Deact 2 data have different parameters from those of Deact 1, the latter agreeing with those of
Actual samples, the Deact 2 data were not considered
further.
Thus
Michel 1 = (0.061 ± 0.002) aca + (0.313 ± 0.20), r2
0.946
Michel 2
(0.054 ± 0.001) aca
(0.043 ± 0.086), r2
0.947
ChE-tel
(4.44 ± 0.30) aca
(5.11 ± 0.99), r2 = 0.959
ChE-tel
(80.2± 5.8) Michel 2 + (0.028 ± 0.240), r2
0.947
ChE-tel (72.0± 9.4) Michel 1 (27.0± 3.0), r2 0.938
The Michel 1/aca and ChE-tel/Michel1 comparisonsshow
non-zero intercepts, indicating
that the Michel 1 data do
require concurrent experimental correction forchange ofpH
in reagentblanks and thatuse ofMichel’stablesisinadequate,at leastover thisconcentrationrange.The ChE-tel/
aca relationship
(Figure IA) appears to be best correlated
below aca values of 15 hUlL (about 60 ChE-tel units). Since
the upper limit of the aca linearity range is 14 hU/L (5), this
may explain why Deact 2 data are not as linear as Deact 1
=
=
=
=
=
=
-
=
=
-
=
=
=
-
=
data.
Data recalculated from an FAA study (8) for blood plasma
led to the expression ChE-tel
110 Michel 2 + 1.38. The
equation given in that study (8) has a slope 1.37 times that
from the present study, has a significantly
non-zero intercept, and no errors are given. In the FAA study a buffer
different
from that recommended was used. Furthermore,
the Michel normal range corresponds to 0.5 to 1.0 units (7).
The ChE-tel normal range (5) is 45 to 90 ChE-tel units.
=
CLINICAL CHEMISTRY, Vol. 30, No. 2, 1984 309
Table 1. Comparison of the Linearity of Each Method for Deactivated Samples Used to Dilute Fresh
Sera (Deact) and for Fresh Sera (Actual), and Mean Data
ApplIcation
Intercept
Slope
Michel 2/aca
Deact I
Deact 2
0.053
0.047
0.054
(0.055 ± 0.002)
(0.054± 0.001)
Actual
Mean line
Mean line8
Michel 1/aca
Deact 1
Deact 2
Actual
Mean line
Mean line8
0.995
0.622
0.933
0.296
(0.351 ± 0.153)
(0.313 ± 0.020)
0.998
±
(4.44
±
0.64)
0.30)
ChE-tel/Michel 1
Deact 1
5.8)
78.2
Deact 2
Actual
Mean line
64.9
Mean line8
Data from Deact I and Actual pooled.
20
5
10
15
20
25
ACTIVITY. k U/L
I
1.00
as:
p
0
5
10
15
ACTIVITY, k U/L
(a&
20
0.980
0.950
0.776
0.947
0.973
0.966
0.954
0.762
0.938
sion,
Wilmington,
DE, 1981.
6. MichelHO. An electrometric method for the determination of
red blood cell and plasma cholinesterase activity. J Lab Clin Med
25
100
A.B.C
80
Key
0P.,ll.iiy
I-1060
UJ
I
40
Fully
Q-F,ssl.
d.I,Il,..d
,.,u,,
C.) 20
0.4
MICHEL
0.8
1.2
2 (pH/h)
Ag. 1. ComparisonofChE-tel/aca
(A), Michef2/ace (,
Michel 2 (C) methods.
and ChE-tet/
the FAA equation, the normal range predicted from
the Michel 2 method for the ChE-tel method would be 56 to
111. Our equation predicts a normal range between 40 to 80.
The normal range in aca units (5) is 7 to 19 kU/L.
From
This study was funded by USPHS Grant ES-00159. Our Lady of
Mercy Hospital, Mariemont,OH-especiallypathologist James
310
0.961
animals. Crit Rev Toxicol 1, 153-202 (1972).
4. Wills JH. Blood cholinesterase:Assay methods and considerations. Lab Manage 20, 53-64 (1982).
5. Du Pont automatic chemical analyzer test methodology for
pseudocholinesterase,Du Pont Company, Clinical
Systems Divi-
B
F.s
0.959
References
1. Vandekar M. Minimizing occupational exposure to pesticides:
Cholinesterase determination and organophosphorus poisoning.
Residue Rev 75, 67-79 (1980).
2. Snow JC. Manual of Anesthesia,
1st ed., Little, Brown, New
York, NY, 1977, p 122.
3. Wills JH. The measurement and significance of changes in the
cholinesterase activities of erythrocytes and plasma in man and
0
1.50
‘S
± 2.82)
± 0.99)
Quinn_are thanked for the useoftheir
Du Pontaca analyzer and
their laboratory. Du Pont-aca Division
isthankedforproviding the
aca reagentpacks, and Pfizer Diagnostics for one of the ChE-tel
kits.
‘F.
()
-(5.11
±
80
40
-(2.78
-24.4
-(19.0
4.2)
-(27.0 ± 3.0)
± 15.9)
± 9.4)
-.100
I
0.993
0.969
0.883
-28.6
-20.2
46.6
(57.7
(72.0
0.987
8.65
-5.57
5.37
(2.15 ± 3.19)
(0.028 ± 0.240)
7.8)
±
±
0.933
0.946
-3.76
-6.79
-6.75
74.8
(65.2
(80.2
0.086)
0.996
0.897
0.947
0.324
86.5
56.6
Actual
Mean line
Mean line8
±
0.954
0.056
(3.87
ChE-tel/Michel 2
Deact 1
Deact 2
_i’
-(0.043
4.71
2.60
4.11
Mean line
Mean line8
I
-0.138
-(0.007 ± 0.014)
0.992
0.060
0.063
(0.063± 0.004)
(0.061± 0.002)
ChE-tel/aca
Deact 1
Dead 2
Actual
0.046
0.262
CLINICAL CHEMISTRY, Vol. 30, No. 2, 1984
34, 1564-1568 (1949).
7. U.S. Dept. of HEW. Criteria for a recommended standard for
occupational exposure during the manufacture and formulation of
pesticides,U.S. DHEW (NIOSH), No. 78-174, 1978.
8. Crane C, Sanders D, Abbott J. Comparison of serum cholinesterass methods II, Federal Aviation Administration,
1972,unpublished report.
9. ChE-tel Manual Reagent Set Procedure for the determination of
cholinesterase. Pfizer Diagnostics Division, Pfizer, Inc., New York,
NY, 1974.
10. Daniel W. Biostatistics: A Foundation
forAnalysis
in the Health
Sciences,2nd ed., Wiley and Sons, New York, NY, 1978.
11. Garry PJ, Routh JI. A micro method for serum cholinesterase.
Clin Chem 11, 91-96 (1965).
12. Ellman GL, Courtney KD, Andres V, Featherstone RM. A new
and rapid colorimetric determination of acetylcholinesterase activity. Biochem Pharmacol 7,88-95(1961).
13. Gal EM, Roth E. Spectrophotometric methods for determination of cholinesterase activity. Clin Chim Acta 2, 316-326 (1957).