CLIN.CHEM. 40/12,2282-2287(1994)
#{149}
Automation
and Analytical
Techniques
Urea and Lactate Determined in 1-pt Whole-Blood Samples with a Miniaturized
Thermal Biosensor
Bin Xie,’ Ulrika
Harborn,
Michael
Mecklenburg,
and Bengt
A miniaturized flow-injected thermal biosensor was developed for the determination of urea and L-lactate in undiluted blood in 1-L samples. The sensor employed a
small enzyme column constructedof stainlesssteeltub-
ing and microbead thermistors. Urease and lactate oxidase/catalase were separately immobilized onto controlled-pore glass beads, which, in turn, were charged into
the enzyme column. With a flow rate of 70 .tL/min, linear
analytical ranges from 0.2 to at least 50 mmol/L and 0.2 to
14 mmol/L were obtained for urea and lactate, respectively. The relative standard deviations (CVs) for measurements of analyte in buffer were 0.91% for urea and
1.84% for lactate. For urea in whole blood, the CV for 50
determinations was 4.1%. Contrived samples containing
various concentrations of urea and L-lactate in whole
blood were determined with this sensor and with a spectrophotometric method. Comparisons of the results gave
correlation coefficients of 0.989 and 0.984 for 30 blood
urea and 30 blood lactate assays inconcentrationsranging from 4 to 20.9 mmol/L and from 1.7 to 12.7 mmol/L
respectively.
Indexing Term.:
immobillzed enzymes/flow-injection analysis
The determination
of such metabolites
as glucose,
lactate, and urea in whole blood is of central importance
in clinical diagnostics.
This information
provides guidelines for determining
our general
state of health. Although routine monitoring
of personal health allows the
detection of disease states at an early stage, such monitoring is expensive and inconvenient.
Development
of a
simple, inexpensive,
and reliable
analyzer
for use in
decentralized
health facilities
is of special interest.
Small sample volumes and small size would be preferred
for this type of instrument.
A great deal of effort has
been devoted to the development
of such devices, especially for glucose determinations
(1-10).
Biosensors
have many advantages,
including convenient operation, rapid response, and low cost. The majority of these systems are based on electrochemical
or
optical detections
(11-17),
which provide good sensitivity and selectivity.
However, their application in wholeblood measurements
is limited because of interference
from electroactive
species, ions, turbidity,
and products.
Moreover, transducers,
such as electrodes,
require
frequent recalibration
for long-term operation.
As an alternative,
we have designed and constructed
Pure and Applied Biochemistry, Lund University, Box 124,
S-22100 Lund, Sweden.
‘Author for correspondence.
Fax mt + 4646104611, E-mail
[email protected].
Received May 2, 1994; accepted September 9, 1994.
2282 CLINICAL CHEMISTRY, Vol. 40, No. 12, 1994
Danielsson
a miniaturized
flow-injected
thermal biosensor based on
the conventional
enzyme thermistor (18, 19). Compared
with other biosensors,
thermal biosensors
are broadly
applicable and are intrinsically
insensitive
to the optical
and electrochemical
properties of the sample. Moreover,
because the thermal transducers
(e.g., thermistors)
are
usually insulated
from the fluid and intrinsically
are
very stable these biosensors
do not require
frequent
recalibration
during long-term operation. The feasibility of using such a system for determining
whole-blood
glucose has previously been demonstrated
with a microcolumn-based
thermal biosensor (20). By reducing the
sample volume, glucose concentrations
20 mmol/L in
whole blood could be analyzed without any pretreatment of the sample. This extended linear range was
attributed
to a relative increase in sample dispersion
(21).
Here, we report the development
of a method for the
determination
of urea and L-lactate in undiluted
wholeblood samples by use of a similar miniaturized
thermal
biosensor.
As in our previous blood glucose studies, we
used 1-jL sample volumes to extend the linearity
of the
analytical
range into the relevant
physiological
concentration range.
Materials and Methods
Materials
U-Lactate oxidase (EC 1.1.3.2; 39 kU/g) from Pediococcus sp., urease (EC 3.5.1.5; 66 kU/g) from jack beans, and
catalase (EC 1.11.1.6; 19 000 kU/g) from beef liver were
purchased from Sigma Chemical Co. (St. Louis, MO). Human blood samples were taken from healthy volunteers.
Analytical grade L-lactate (sodium salt; Sigma) and urea
(E. Merck; Darmstadt,
Germany) were used as the standards; all other reagents
were analytical grade. The test
kits for the comparative
measurements
of blood urea
(Urea Granutest
15 Plus; Diagnostica
Merck, Darmstadt,
Germany)
and lactate (LACT MPR 1; Boehringer
Maimhelm, Mannheim,
Germany) were supplied by the manufacturers.
Propylamino-derivatized
controlled-pore
glass
beads (CPG; particle diameter 125-140 .tm, pore diameter
50 nm) was obtained from Steinachglas,
Steinach, Germany. The micro-bead thermistors
(K19120%112K) used as
the temperature
transducer
were purchased
from Siemens, M#{252}nchen,
Germany.
Preparationof the Enzyme Matrixes
All enzymes used in the study were covalently immobilized on spherical CPG beads as follows: The propylamino-CPG
was activated with 25 milL glutaraldehyde
in 0.1 molIL sodium phosphate buffer, pH 7.0, at room
temperature
for 1 h under reduced pressure, washed
exhaustively
on a B#{252}chner
funnel with distified
water,
and finally stored in water at 4#{176}C.
Lactate oxidase (-50
U) and catalase (>20 000 U) were added to 50mg of wet,
activated
CPG suspended
in 500 L of buffer. The coupling was allowed to proceed at 4#{176}C
overnight
on a
shaker. Before packing it into a column, we thoroughly
washed the enzyme preparation with phosphate buffer,
then packed -25 L of the matrix into each column.
Similar procedures
were used for urease immobilization, adding 800 U of urease to 0.5 g of wet, activated
CPG suspended in 0.5 mL of buffer.
Preparationof Samples
The L-lactate samples were dissolved
in a buffer containing
137 mmol/L NaCl, 2 mmolIL EDTA, and 100
minol/L
sodium phosphate (pH 7.0). Urea samples were
prepared
in the same buffer but also containing
2
mmol/L
-mercaptoethanol.
Urea and lactate calibrators were prepared in their respective buffers. Venous
blood samples
were collected
into EDTA-containing
tubes and stored on ice until the measurements
were
performed.
Biosensors
The sensor device, 24 mm (o.d.) x 54 mm, was composed of three parts: a cylinder, and two cover mounts
that could be easily disassembled
for convenient access
to the enzyme column (Fig. 1). The enzyme column,
15
mm x 1.5 mm (i.d.)/1.7 mm (o.d.) was constructed
of
stainless-steel
tubing and placed in the middle of the
cylinder.
A thin copper foil (0.1 mm thick), onto which a
stainless-steel
tubing [0.3 mm (o.d.)] was tightly and
densely coiled, surrounded
the column. This foil served
as an adiabatic
shield by preserving
the heat of the
outlet stream. The adiabatic layer was insulated
by a
2-mm-thick
layer of Plexiglas,
which in turn was covered with an aluminum
jacket, which functioned
as a
heat sink. Stainless-steel
tubing [0.4 mm (o.d.)1 was
wound in parallel coils around the inside surface of the
jacket to equilibrate
the temperature
of the incoming
fluid. The microbead
thermistors
were mounted
onto
gold capillary tubings [0.3 mm (o.d.)1 with heat-conductReference
Pre-heating
Enzyme
Measurement
thermistor
coil
column
thermistor
\
/
Plexiglass
jacket
I
/
Adiabatic
coil
I
/
I
Aluminum
heat sink
Aluminum
side-cover
ing epoxy. These gold tubings together with the thermistors functioned as temperature-sensitive
elements and
were positioned close to the enzyme column. The entire
aluminum
cylinder and the two cover mounts were then
enclosed by a 3-mm-thick
layer of Plexiglas.
ThermometricAssay
The setup for the thermometric
assay consisted of a
peristaltic pump (Alitea pump unit C-4V; Ventur Tekniska AB, Uttran,
Sweden), an HPLC valve suitable for
1-L samples (Type C14W; VIGI AG, Valco Europe,
Schenkon,
Switzerland),
a preheat sink, the sensor device, a dc-coupled Wheatstone
bridge equipped
with a
chopper-stabffized
amplifier (constructed
at our institute), and a chart recorder.
During the operation, the
sensor was kept in an aluminum
box insulated
with
polyurethane
foam to minimize
interferences
from
changes in the environmental
temperature.
The effect
on the thermal signal detection
was insignificant
for a
change of about ±2#{176}C
at room temperature
(normally
22#{176}C),
even without temperature
control. Because the
sample volume was only 1 L in an assay run at a flow
rate of 70 tL/min, there was good margin for temperature equilibration
before the sample reached the enzyme column via the heat-sink
inside the aluminum
box. The temperature
change corresponding
to the enzyme reaction (usually
in the range
of 102-10#{176}C)
taking place in the column was registered
with the
Wheatstone
bridge. At maximum sensitivity,
the signal
changes 100 mV per 10#{176}C.
ComparisonAssays
For comparison, we also used methods based on ultraviolet spectrophotometric
detection
to determine
the
urea and U-lactate concentrations
in whole blood.
With the Urea Granutest 15 Plus test kit, the plasma
required for the assay was prepared by centrifuging
the
blood specimen for -5 mm at 1000g. The determination
is based on the reactions catalyzed by urease and glutamate dehydrogenase
(EC 1.4.1.3), with the accompanying consumption
of NADH being detected spectrophotometrically
at 340 nm (22). These determinations
were
assumed to correspond to the concentration
of urea in
whole blood, there being only slight differences in urea
concentration
between plasma and whole blood (23).
In the LACT MPR 1 test kit, lactate plus NAD
were
converted to NADH and pyruvate by lactate dehydrogenase (EC 1.1.1.27); the pyruvate
in turn was broken
down by alanine
aminotransferase
(EC 2.6.1.2). The
NAI)H produced was measured at 340 nm (24). Before
the assay the sample was deproteinized
by mixing
1
molIL perchioric acid with an equal volume of blood and
centrifuging;
the supernate was used for the determination. The results obtained
with this method correspond
to those for whole blood, according to the manufacturer’s
calibration
procedure.
Principles
25mm
-5
Fig. 1. Schematic diagram of the sensor construction.
According to the urease reaction (25), urea is hydrolyzed into the products below by urease in neutral phosphate buffer:
CLINICAL CHEMISTRY, Vol. 40, No. 12, 1994
2283
3H20
+
(NH2)2C0
Urease
2NH4
+
+
HC0
0H
+
3
H
+
C-)
0
where tH is the enthalpy change of the reaction.
For the lactate assay, lactate oxidase and catalase
used to convert U-lactate:
are
a
a)
0)
C
Cs
U-Lactate
+02
Lactate
oxidase
‘pyruvate
+
H2O2
+
iXH,
Catalase
H20
+ ‘/202+
C)
oi
E
zH2
0
where H1 and iH2 are the enthalpy changes produced
by the lactate oxidase and catalase reactions,
respectively. This reaction
scheme provides additional
heat
generated by the reduction of H202 (100 kJ per 1 molIL)
and by the partial regeneration
of 02. This extends
the
linear range
of the lactate assay and eliminates
the
presence of H202, which could otherwise affect the stability of the immobilized
enzymes (26).
and DiscussIon
Propertiesof the Sensor
Results
The linear range and the operational
stability of the
sensor were evaluated
with lactate and urea calibrators
in buffer. As mentioned above, the hydrolysis of urea is
limited only by the amount of enzyme in the column. In
contrast, the lactate assay depends also on the amount
of dissolved oxygen. In this study, we used the 1-tL
samples to extend the linear range of the lactate analysis.
The thermal calibration curves for the urea and U-lactate calibrators
were used to evaluate the results from
whole-blood
measurements
over the ranges of 0.5-50
mmol/L
for urea and 0.5-20 mmol/L for L-lactate.
For
urea a broad linear range at least 50 mmol/L was obtained (Fig. 2, top). As the results indicate, the enzyme
columns had adequate catalytic
capacity
for full conversion of the corresponding
substrates
even though the
amount of enzymes
(26 tL column volume) used was
much less than that (>200 pi) in the conventional
enzyme thermistors.
The linear range of U-lactate analyses extended
only to 14 mmoIIL (Fig. 2, bottom), limited by the amount of dissolved oxygen available. Nevertheless,
this concentration
range is adequate
to
measure
lactate concentrations
typically
found in the
blood, the normal physiological
range being 0.55-2.22
mmol/L
(5-20 mg/100 mL) for lactate [for urea, it is
1.33-4.33
mmol/L (8-26 mg/100 mL)1 (27).
During exercise, however, the concentration
of lactate
in blood often exceeds 14 mmol/L (28). In similar analytical systems, samples with extremely high concentrations of analyte can be assayed by further
reducing
the
sample volume (20, 21) or by sample dilution. In addition, we found that the linear range and the sensitivity
of lactate
assay were significantly
affected by the
amount of immobilized
catalase
and lactate
oxidase.
Increasing
the amounts
of immobilized
lactate oxidase
2284
CUNICAL CHEMISTRY, Vol.40,No.12,1994
0
10
20
30
Urea,
mmol/L
40
50
3
00
0
i-
2
a,
0)
C
Ce
C.)
ci.
E
I-
0
Lactate,
mmol/L
Fig. 2. The linear range of the urea samples results (top) and the
L-Iactate determinations (bottom) in buffer made with the thermal
biosensor.
the peak height
and decreased
the linear
range, an indication that the oxygen diffusion rate in
the buffer limited the reaction rate. On the other hand,
decreasing the amount of immobilized catalase resulted
in a decrease
in both the linear range and the peak
height. This insufficient conversion of the hydrogen peroxide by catalase in turn reduced the oxygen supply and
decreased the total thermal
output.
The lower detection
limit (signal-to-noise
ratio
2)
was 0.1 mmoIIL for both analytes in 1-tL samples. Increasing the sample volume lowered the detection limit
but only at the cost of decreasing the linear range.
The linearity and sensitivity
of flow-injected
biosensors are affected by the flow rate. This is particularly
true for the extensively
miniaturized
thermal biosensors since the enzymatic catalysis and the thermal detection are time dependent. Too high a flow rate reduces
the sample residence time, which in turn reduces the
substrate
conversion
efficiency; consequently,
the thermal response is decreased. On the other hand, reducing
the flow rate resulted in poor linearity because of increased heat leakage (21). In this system, we found the
increased
optimal
flow rate to be -70 /11.1mm; this allowed a
sampling
rate of 40 samples per hour.
The operational
stability
of the biosensor was evaluated by performing
>100 manual injections of calibrator
solutions (50 mmol/L urea and 10 mmol/L L-lactate) for
each enzyme column. The relative standard
deviations
(CVs) were 0.91% for the urea assay and 1.84% for the
L-lactate assay. Comparison
of the measurement
series
showed that, at the same concentration,
the L-lactate
response is about three time that of urea. This result is
consistent with other studies (18, 29) in reporting enthalpies of -180 kJ per mol/L for lactate catalyzed
by
lactate oxidase/catalase
and 61 kJ per mol/L for urea
catalyzed by urease in phosphate buffer.
ers. Fig. 3 shows the different responses. As can be seen,
this nonspecific heat was related to the concentration
of
the salts. The thermal response was affected more by the
citrate buffer than by the phosphate
buffer at equivalent
salt concentrations,
and the effect was more pronounced
with KC1 than with NaC1. The relative effect of these
salts decreased
as the sample concentration
increased.
In whole-blood measurements,
the low concentration
of
KC1 (-3.7 mmol/L) had an insignificant
effect. However,
given the considerably
higher concentration
of NaCI
(-137 mmol/L) in whole blood, it was necessary
to supplement the buffer with NaC1 to eliminate
nonspecific
heat effects for injection of whole-blood samples.
Assay of Blood Samples
FactorsAffectingthe Assay
The determination
of urea and lactate are affected by
the enzyme activity and the transducer
stability.
Because the thermistors
were very stable and isolated
from the buffer solution, the most important
factors
affecting the determinations
were those that influenced
the enzyme reaction,
such as enzyme inhibitors,
pH
changes, buffer composition,
and nonspecific heat.
The most common inhibitors
of urease are heavy
metal ions, and some metals can inactivate the enzyme
at very low concentrations.
The effect could not be
avoided in this flow system, given the use of metal
tubing in the heat sink and the inlet connections;
moreover, whole blood itself contains free metal ions. Therefore, we included EDTA to reduce the inhibitory
effects
of the metal ions. Oxidants can also affect urease activity, particularly
in long-term operation. As we found
previously
(30), the antioxidant
/3-mercaptoethanol
stabilizes urease. Similar results were found in this study
with use of phosphate
buffer containing
2 mmol/L
EDTA, 2 mmoIJL /3-mercaptoethanol,
and 137 mmol/L
NaCl (not shown). We also investigated
the effect of
other compounds-e.g.,
HCO
(10-50 mmol/L), NaCl,
and KC1-on
the urease reaction. Including HCO did
not affect the thermal
response,
but NaCl (>50 mmol/L)
and KC1 (>5 mmol/L) did. These ion effects are discussed below.
The effect of pH on the lactate oxidase reaction was
studied
with two different buffer systems, phosphate
buffer (pH 6-8) and citrate buffer (pH 5-6). In the experiment, the same buffer was used as running
eluent
and to prepare the L-lactate calibrators (5-20 mmolJL).
The calibration
curves obtained with these two buffer
systems
showed
no significant
differences
between
them. Furthermore,
no significant
interference
was observed from HCO (2-100 mmol/L) or other metabolites,
such as ascorbic
acid (1-6.8 mmolIL), pyruvate (0.050.7 mmol/L), and oxalate (0.1-1.5 mmol/L), in phosphate
buffer, pH 7.0. However, the peak height increased
because of the heat of solvation when lactate calibrators
containing
KCI or NaC1 were injected. To ascertain the
relative contribution
from these salts on the thermal
signal, we analyzed
several samples containing various
amounts of NaCl (10-500
mmol/L) and KCI (10-500
mmol/L) in citrate (pH 6.0) and phosphate (pH 7.0) buff-
Human venous whole-blood
samples were stored on
ice after collection to stabilize the concentration
of the
analytes. This is particularly
important for lactate determinations,
because its concentration
increases
dramatically after being withdrawn from the body. For the
comparison studies, we divided each blood sample into
two portions and immediately
analyzed them with the
biosensor and the spectrophotometric
method.
Urea in whole blood was determined
in 0.1 mol/L
phosphate buffer (pH 7.0) containing 2 mmolIL EDTA, 2
mmol/L /3-mercaptoethanol,
and 137 mmolIL sodium
chloride. Blood samples containing
various concentrations of urea were prepared by adding 1-15 tL of a 200
mmoIJL urea calibrator to 200-1LL aliquots from a single
whole-blood specimen. The urea concentration
in whole
blood was determined
by comparing
the thermal response from these blood samples with the calibration
curve established
from urea buffer calibrators. We then
compared these results with the blood urea sample determinations
obtained
with the spectrophotometric
method (Fig. 4). The correlation between the two methods was r = 0.989 for 30 blood samples ranging in
concentration
from 4 to 20.9 mmol/L urea. The standard
1,0
C)
0
0,8
C’)
0
a)
0,6
0)
C
CS
.=
C)
a.
E
I!
0,4
0,2
0,0
0
100
200
300
400
500
cr concentration,
mmo/L
Fig. 3. Contributionof KCI and NaCl to nonspecific thermal responses of the lactate sensor in citrate (pH 6) and phosphate buffers
(pH 7), respectively.
A, KCI in citrate buffer; B.NaCIin citrate buffer C, KCIinphosphatebuffer;and
D, NaCI in phosphatebuffer.
CLINICAL CHEMISTRY, Vol. 40, No. 12,
1994
2285
surements
were compared with the calibration curve for
the lactate calibrators
in buffer and with the results of
the comparison
method.
A correlation
coefficient
of
0.984 was obtained
between these two methods for 30
lactate assays ranging
in concentration
from 1.7 to 12.7
mmol/L (Fig. 4). In this study, the standard
deviations
for the slope (Sb) and the intercept (Sa) were 0.0320 and
0.2528, respectively.
The 95% confidence limits for the
intercept
and slope were: a = 0.0881 ± 0.5157 and b =
0.92827 ± 0.06528. No systematic
errors were observed
in the comparison
study, although others have reported
a gradual, time-dependent
increase
in the lactate concentration
in blood samples
(31). This drift in lactate
concentration
in blood samples makes the performance
of reproducibility
studies difficult to interpret.
However,
we find it worth noting that our method correlates
well
with the LACT MPR 1 test kit.
1
0(0
Ec
Blood urea, mmol/L
with the UV method
0-
Eo
E
0
Fig.
5
10
15
Blood lactate, mmol/L
with the UV method
4.Correlation
betweentheblood urea measurements (top) and
the blood lactate concentrations (bottom) measured by the thermal
biosensor and bythe spectrophotometric
method.
Shown are results for 30 blood samples ranging in urea content from 4 to 20.9
mmoVL and 30 blood samples ranging in lactate concentration from 1.7to 12.7
mmoUL
deviations
for the slope (Sb) and the intercept (Sa) were
0.0267 and 0.3581, respectively.
The use of the appropriate t-value for 28 degrees
of freedom (t = 2.04) gave
the 95% confidence limits for the intercept
and slope as
a = 0.4036 ± 0.7305 and b = 0.96324 ± 0.05448. The CV
was 4.1% for 50 determinations
made on the same day of
one blood sample
containing
a mean value of 0.37
mmol/L urea. We used the same column for blood urea
determinations
for about a week at room temperature
without significant
decrease
in the thermal response.
The analysis
of additional samples should be possible,
given that no clogging in the column was observed after
50 whole-blood sample injections.
The determination
of whole-blood lactate was carried
out similarly.
The buffer used in these analyses
consisted of 2 mmol/L EDTA and 137 mmolJL NaCl in 0.1
mol/L phosphate
buffer, pH 7.0. Blood-based
lactate calibrators were prepared as described
for the urea calibrators. The results obtained from these blood sample mea2288 CLINICALCHEMISTRY, Vol. 40, No. 12, 1994
In conclusion, we have demonstrated
the feasibility
of
developing a universal biosensor for determining
wholeblood metabolites
by measuring
blood urea and lactate
in a miniaturized
flow-injected
thermal
biosensor. The
use of small sample volumes and a spherical,
rigid enzyme support
material
with uniform
particle
size allowed virtually
all of the blood cells to pass through the
enzyme column. No clogging was observed for the number of samples (up to 100) studied here, irrespective
of
hematocrit.
The device requires only occasional calibration, and the universality
of the method makes it simple
to implement.
The simple reaction scheme in this system does not require
any special reagents
or chromophores that can cause interference.
Therefore,
the
method shows promise
for determining
many
other
blood metabolites
besides glucose, urea, and lactate. The
linear range of the oxidase reactions, exemplified by the
lactate oxidase reaction described here, was extended by
using very low sample volumes. Alternatively,
it might
be possible to extend the linear
range
of the oxidase
reaction
by electrochemically
regenerating
an electron
mediator,
such as ferrocene,
to replace oxygen in the
buffer (32).
The approach described
here could easily be applied to
the simultaneous
analysis
of multiple analytes
(30).
The use of immobilized
enzyme columns provides
an
economical
alternative
to single test kits because each
column can be used for at least 100 determinations.
In
the future, reliable, inexpensive,
and convenient
portable devices based on this principle could be developed for
decentralized
and home health monitoring.
This work was supported by a grant from the National
Board for Technical Development (NUTEK).
Swedish
References
1. Brooks KE, Rawal N, Henderson AR. Laboratory assessment of
three new monitors of blood glucose: Accu-Chek II, Glucometer H,
and Glucoscan 2000. Clin Chem 1986;32:2195-200.
2. Matthews DR. Pen-sized digital 30-second blood glucose meter.
Lancet 1987;i:778-9.
3. Kiser EJ, Rice EG, Tomasco MF. Blood separation and analyte
detection techniques and test strips. Eur Pat 415679A2. May 1991.
4. Kost GJ, Wiese DA, Bowen TP. New whole blood methods and
instruments: glucose measurement
and test menus for critical
care. J mt Fed Clin Chem 1991;3:160.-72.
5. Groenlund
B, Groenlund
L, Macisbad S, Fogh-Anderson
N.
Clinical evaluation of a prototype glucose electrode. Biosens Bioelectron 1992;7:683-7.
6. Cvorisces D, Petroveck M, Cunovic M, Stavljenic A. Evaluation
of the Abbott EPx clinical chemistry analyzer. Kiln Labor 1992;
38:544-8.
7. Nankai S, Kawaguri M, Yoshioka T, Tsutsumi H, Terao K,
Tanimoto N, et al. Quantitative
analysis method and its system
using a disposable sensor. Eur Pat 471986A2. February 1992.
8. Gluhak J, Topic E, Pavicic V. A fast method for glucose
determination
in blood hemolysate
on the Technicon
RA-1000.
Laboratoriumsmedizin
1992;16:40-2.
9. Schembri CT, Ostoich V, Lingane PJ, Burd TL, Buhl SN.
Portable simultaneous multiple analyte whole blood analyzer for
point-of-care testing. Clin Chem 1992;38:1665-70.
10. Ashworth L, Gibb I, Alberti KGMM. HemoCue: evaluation of
a portable photometric system for determining glucose in whole
blood. Clin Chem 1992;38:1479-82.
11. Mizutani F, Sasaki K, Shimura Y. Sequential determination
L-lactate and lactate dehydrogenase
with immobilized enzyme
electrode. Anal Chem 1983;55:35-8.
12. Mascini M, Moscone D, Palleschi G. A lactate electrode with
lactate oxidase immobilized on nylon net for blood serum samples
in flow systems. Anal Chim Acta 1984;157:45-51.
13. Wilson GS, Zhang Y, Reach G, Moatti-Sirat
D, Poitout V,
Th#{233}venot
DR, et al. Progress toward the development of an
implantable
sensor for glucose. Clin Chem 1992;38:1613-7.
14. Hkanson
H, Kyrolainen
M, Mattiasson
B. Portable system
for continuous ex vivo measurements
of lactate. Biosens Bioelectron 1993;8:213-7.
15. Girotti S, Grigolo B, Fern E, Ghini S, Carrea G, Bovara R, et
al. Bioluminescent
flow sensor for the determination of L-(+)lactate. Analyst 1990;115:889-94.
16. Dremel BAA, Li SY, Schmid RD. On-line determination of
glucose and lactate concentrations in animal cell culture based on
fibre optic detection of oxygen in flow-injection analysis. Biosens
Bioelectron 1992;7:133-9.
17. Wolibeis OS, Li H. Fluorescence optical urea biosensor withan
ammomum optrode as transducer.
Biosens Bioelectron
1993;8:
161-6.
18. Danielsson B, Mosbach K. Theory and application of caloriof
metric sensors. In: Turner APF, Karube I, Wilson GS, eds. Biosensore: fundamentals
and applications. London: Oxford University
Press, 1987:575-95.
19. Danielsson B. Calorimetric
biosensors. J Biotechnol 1990;15:
187-200.
20. Xie B, Hedberg U, Mecklenburg M, Danielsson B. Fast determination of whole blood glucose with a calorimetric micro-biosensor. Sens Actuat 1993;15:141-4.
21. Xie B, Winquist F, Danielsson B. Miniaturized
thermal biosensors. Sens Actuat 1993;16:443-7.
22. Talke H, Schubert GE. Enzymatische
harnstoflbestimmung
in
blut und serum im optiachen test nach warburg. Kiln Wochenschr
1965;43:174-5.
23. Sunderman FW, Boerner
F, eds. Normal values in clinical
medicine. Philadelphia:
Saunders, l95O:845pp.
24. Noll F. L-(+)-Lactate
determination
with LDH, GPT and
NAD. In: Bergmeyer HU, ed. Methods of enzymatic analysis, 2nd
ed. New York and London: Verlag Chemie Weinheim and Academic Press, l974:2302pp.
25. Jespersen ND. A thermochemical
study of the hydrolysis of
urea by urease. J Am Chem Soc 1975;97:1662-7.
26. Guilbault GG, Danielsson
B, Mandemus
CF, Mosbach K.
Enzyme electrode and thermistor
probes for determination
of
alcohols with alcohol oxidase. Anal Chem 1983;54:1582-5.
27. Henry RJ, Cannon DC, Winkelman
JW, eds. Clinical chemistry: principles and technics. Hagerstown,
MD: Harper & Row
Publishers, l974:l629pp.
28. Shimojo N, Fujino K, Kitahashi 5, Nakao M, Naka K, Okuda
K. Lactate analyzer with continuous blood sampling for monitoring blood lactate during physical exercise. Clin Chem 1991;37:
1978-80.
29. Scheller
F, Siegbahn N, Danielsson
B, Mosbach K. Highsensitivity enzyme thermistor determination of L-lactate by substrate recycling. Anal Chem 1985;57:1740-3.
30. Xie B, Mecklenburg M, Danielsson B, Ohman 0, Norlin P,
Winquist F. Development of an integrated thermal biosensor for
the simultaneous
determination of multiple analytes. Analyst
1994; in press.
31. Savory J, Kaplan A. A gas chromatographic
method for the
determination of lactic acid in blood. Clin Chem 1966;12:559-69.
32. Xie B, Khayyami M, Nwosu T, Larsson P0, Danielsson B.
Ferrocene-mediated
thermal biosensor. Analyst 1993;118:845-8.
CLINICALCHEMISTRY, Vol. 40, No. 12,
1994
2287