131
CIjnica Chimica Acta, 69 (1976) 131-135 @ Elsevier Scientific Publishing Company,
Amsterdam
- Printed
in The Netherlan‘ds
CCA 7753
SHORT COMMUNICATION A RAPID METHOD FOR THE MEASUREMENT OF PLASMA GLUCOSE CONCENTRATION
W.A. TELFER Department
(Received
BRUNTON
of Clinical
December
and I.W. PERCY-ROBB
Chemistry,
The Royal
Infirmary,
Edinburgh,
EH3
9YW
(U.K.)
19, 1975)
Summary A modification of the manual glucose oxidase-gum guaiacum method of Shipton, B., Wood, P.J. and Marks, V. ((1975) Med. Lab. Tech. 32, 33), for use with plasma, is described. The method enables rapid measurement of plasma or serum glucose concentration between 1 and 22 pmol/l using 10 ~.tl of sample and a single colour reagent. Experience of the method in routine use has shown that the coefficient of variation, between batch, is 2.1% and the time taken for analysis of 1 sample, standards and a quality control sample is 12 min (including centrifugation of blood to obtain plasma).
Introduction The glucose oxidase-peroxidase-o-tolidine method of Marks [l] for measuring blood glucose concentration is used in many laboratories for emergency work. The method is however time-consuming and tedious, and has the disadvantage of using o-tolidine, a carcinogen, as the chromogenic oxygen acceptor. More recently Morley et al. [2] and Shipton et al. [3] have described glucoseoxidase methods using gum guaiacum, which has not been implicated as a carcinogen, as the oxygen acceptor. As there is now a body of opinion supporting the use of plasma rather than whole blood for glucose measurement [4--81 we have adapted the method of Shipton et al. [3] for use with plasma and simplified the technique by using a single stable colour reagent. Materials and methods Stock guaiacum solution 3 g gum guaiacum (Hughes
and Hughes,
London)
was added
to 48 ml of
132
ethanol, shaken for 1 h and filtered through Whatman No. 1 filter paper (W. and R. Balston Ltd., Maidstone, England). 80 ml of Triton Xl00 (B.D.H. Chemicals Ltd., Poole, England) were added to the filtrate, and the mixture diluted to 4 1 with acetate buffer (0.5 mol/l, pH 6.2). Working colour reagent 9.9 ml of glucose oxidase (Fermcozyme, Hughes and Hughes, London) and 9.9 mg of peroxidase (Activity RZ = 1, Hughes and Hughes, London) were added to 300 ml of acetate buffer (0.5 mol/l, pH 6.2), mixed and then added to 600 ml of stock guaiacum solution. The solution was left for 2-3 h in the dark before use. Maximum storage at 4°C is 3 weeks. Standards 1. Aqueous. Stock D-glucose solution (100 mmoljl) in saturated aqueous benzoic acid solution. Working standards (1.0, 5.0, 15 mmol/l) were made up by dilution of stock with saturated aqueous benzoic acid solution. 2. Plasma. Pooled human plasma was dialysed against sodium chloride solution (150 mmol/l) until no glucose was detectable (AutoAnalyzer (Technicon Instruments Co., Ltd., Basingstoke, England) and manual glucose-oxidase-otolidine methods). Stock D-glucose solution was then added to give a final concentration of 2.7 mmol/l. “Five minute manual guaiacum method (5MGM)” Blood samples were centrifuged for 5 min at 1000 X g in a bench centrifuge and plasma carefully removed into 5ml plastic tubes. 5 ml of working colour reagent was pipetted into lo-ml plastic tubes, one for each test, quality control specimen and standard, plus a blank. At timed intervals, 10 ~1 of the standards, test samples or controls were added to the colour reagent using a graduated 50 ~1 Hamilton Syringe (V.A. Howe and Co. Ltd., London) and mixed by inversion. After 5 min incubation at room temperature the absorbance was measured at 625 nm in an EEL Spectra spectrophotometer. A u toA nalyzer method The method of Morley et al. [2] was used except reagent was prepared as described above.
that the working
colour
Results The absorption spectrum of the developed colour showed a single broad peak with a maximum absorbance at approximately 610 nm. A wavelength of 625 nm was chosen for these studies since the reduced sensitivity (about 20% reduction) at that wavelength allowed a useful range of glucose concentrations to be measured using the 10 ~1 sample volume. The relationship between colour development and time was examined (Fig. 1). A 5-min incubation was found to produce a satisfactory end-point at each of the concentrations tested. Linearity was determined using D-glucose standards and also with dialysed, pooled plasma to which known amounts of glucose had been added. Between
133
COLOUR DEVELOPMENT 0.7
CURVES
FOR 5MGf.i
1
3
I
7
5
TIME IN M/N Fig.
1. Relationship
between
colour
development
and time.
1 mmol/l and 22 mmol/l colour development was found to be linear. A comparison of the 5MGM was made with the AutoAnalyzer method using plasma samples preserved with sodium fluoride (47 mmol/l). The results are shown in Fig. 2. Accuracy and precision The recovery of glucose added to plasma samples was 99-103% (Table I). One Wellcome Quality Control sample and one sample of dialysed plasma containing D-glucose were included with each batch of analyses. The quoted value for the Wellcome Quality Control sample using manual glucose oxidase methods was 5.94 mmol/l. The found mean value for 19 assays was 5.83 mmol/l and the
5MGM
Glucose
Autoordyzer cwretatron Number
01
3 Plasma
Fig.
2.
0.996x
Conparison -
0.25.
of
concn. fhvnol/l)
5MGM
coefficient ofpain
and
the
J-025
Ir I = 0.99
(n) = 67
15
10 glucose
Cmcn.=IO.996x
Glucose Cmcn
20
Autoanolyzer AutoAnalyzer
method.
- - - - - -, 45”
line;
----,
regression
line
y =
134 TABLE
I
RECOVERY
EXPERIMENTS
-~ Initial
Glucose
Expected
Found
glucose
added
(mmol/l)
mean
concn.
(mmol/l)
Recovery
Range
(%)
(mmol/l)
(mmol/l) 0
2.11
2.77
2.77
2.63-2.83
100
2.17
5.55
8.32
8.55
8.53-8.65
103
8.55
5.0
13.55
13.50
TABLE
99
13.33-13.59
II
COMPARISON
OF
MANUAL
GLUCOSE
METHODS
Glucose
oxidase-gum
guaiacum
Glucose
methods
Marks Shipton Sample Sample
volume
Approx.
time
(1 sample,
for
procedure
1 control
Working
reagents
Number
of pipettings
Carcinogenic
* Approx. glucose,
*
[II
5MGM
Blood
Plasma
Blood/plasma
10 I.rl
10 /.ll
100
7-l
2 min
12 min
Pl
30
min
(plasma)
+ standard)
required
reagents
time
et al. [ 31
oxidase-
u-tolidine
for
subtract
measuring
2
1
3
2
None
None
plasma
glucose,
including
5 min
4 6 o Tolidine
centrfugation.
(For
measuring
blood
5 min).
coefficient of variation was 2.1% (6 operators). The final glucose concentration after adding glucose to dialyzed plasma was 2.77 mmol/l. The found mean value for 14 assays was 2.77 mmol/l with a C.V. of 3%. Within batch precision was measured by running one sample 36 times (4 batches containing 9 samples each). The coefficient of variation within each batch was 0.7%. Interference and inhibition Neither urate up to 0.95 mmol/l nor bilirubin up to 100 pmol/l had any effect upon values obtained with the 5MGM (recovery 100% in both cases). The effects of fluoride ions were examined since all of the blood samples received in our laboratory are preserved in this way. Recovery after addition of fluoride ions up to 95 mmol/l (twice the concentration normally used) was 99% (measured at 2.77 and 8.3 mmol/l). Discussion The development of the 5MGM, and the rapid guaiacum method et al. [3] both arose from the need to introduce a rapid, accurate method for measuring plasma or blood glucose concentration under conditions using a method which excluded the use of carcinogenic
of Shipton and simple emergency reagents. In
135
addition, the method described here does not require a protein precipitation step. In the laboratory, and particularly where manual methods are c,oncerned, carcinogenic compounds should be replaced by non-toxic alternatives. Many of the chromogens which have been used as oxygen acceptors in glucose oxidase methods are now known, or suspected, to be carcinogenic. In particular, derivatives of benzidine, including o-dianisidine and o-tolidine have been specifically included in the 1967 Carcinogenic Substances Regulations, No. 879 and it is therefore important that efforts should be made to exclude these compounds from the laboratory. Replacement of o-tolidine by gum guaiacum or by any other non-carcinogenic chromogen can therefore be regarded as an important positive step in promoting the safety of laboratory staff at work. It is desirable that methods used for providing emergency results should correlate well with the methods normally used in the laboratory for those assays. The results obtained from the 5MGM correlate well with the AutoAnalyzer method (Fig. 2) and accuracy and precision figures compare favourably with those of Shipton et al. [3] for their manual guaiacum method (recovery 100.7X, C.V. 2.5%) and with those of Morin and Prox [9] for their glucose oxidase method using pdiphenylamine sulphonate (recovery 99%, C.V. 3.3%). Comparison of the methods (Table II) shows that the guaiacum methods are quicker and simpler to perform than the o-tolidine method. The 5MGM also has the advantage of using a single colour reagent. The main weakness of the manual guaiacum methods is that it can be difficult to measure small volumes (10 ~1 plasma) reproducibly. Our own experience using a Hamilton Syringe, and studies carried out by Shipton et al. [ 31, indicate however that, a coefficient of variation of 1% can be achieved (within batch). The ability to obtain reliable measurements with sample volumes as small as 10 ~1 is especially useful for paediatric work. References 1 2 3 4 5 6 7
Marks, V. (1959) Clin. Chim. Acta 4, 395 Morley. G.. Dawson. A. and Marks, V. (1968) Proc. Assoc. Clin. Biochem. 5, 42 Shipton, B., Wood, P.J. and Marks. V. (1975) Med. Lab. Tech. 32. 33 Ingram, P.. Ingram. S.. Turtle, S., Sturrock, S. and Applegarth, D. (1971) Clin. Biochem. 4. 297 Morrison. B. and Fleck, A. (1973) Clin. Chim. Acta 45, 293 Henry, R.J. (1974) Clinical Chemistry. Principles and Techniques. Harper and Row. New York Standardisation of the oral glucose tolerance test, Report of the Committee on Statistics American Diabetes Association, June 14, 1968 in: (1969) Diabetes 18, 299 8 Holtkamp. H.C., Verhoef, N.J. and Leijnse. B. (1975) Clin. Chim. Acta 59, 41 9 Morin, L.G. and Prox, J. (1973) Clin. Chem. 19. 959
of the