Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~~ 1875 2
METHOD FOR THE DETERMINATION
OF HEMOGLOBIN ADDUCTS
Background of the Invention
The level of glycation of certain circulating
proteins, e.g. hemoglobin, can be used to monitor the
average blood glucose because glycation is a non-
10 enzymatic, slow and continuous reaction that is prima-
rily dependent on the ambient glucose concentration to
which the hemoglobin is exposed during its residence
time in the circulatory system. These two factors,
glucose concentration and residence time, translate in
15 vivo to the degree and duration of increased blood
glucose concentration (hyperglycemia). Thus, when the
blood glucose level is elevated, as it is in diabetic
people whose diabetes is not well controlled, increased
amounts of glycated hemoglobin are formed. The amount
20 of glycated hemoglobin in an individual's blood reflects
the average blood glucose concentration to which hemo-
globin has been exposed during its life in the circula-
tory system. This period is about 100 days, so the
determination of the concentration of glycated hemo-
MSE #1838.1
- 2 - X11875 2
globin can provide an historical picture of the indi-
vidual's blood glucose profile.
Methods described for the measurement of glycated
hemoglobin include chromatography on an ion exchange
5 column or boronate affinity columns, HPLC and agarose
gel electrophoresis. Each of these techniques has
drawbacks with respect to complexity, costly instru-
mentation, accuracy or variability.
Immuno techniques are available whereby monoclonal
10 antibodies have been developed to react with epitopes
of the glycated hemoglobin, i.e. that derivative of
hemoglobin formed by the nonenzymatic reaction of
glucose with reactive amine groups on the hemoglobin
protein, and facilitate the determination of glycated
15 hemoglobin concentration by standard immunochemistry
techniques such as ELISA or a latex agglutination
assay. Such an agglutination assay is disclosed in
U.S. Patent 4,970,171 wherein there is described an
immunoassay for glycated hemoglobin, e.g. HbAlc, a
20 variety of hemoglobin with glycation at the ~i subunit
in a blood sample, which involves the steps of:
a) treating the blood sample with a thiocyanate
salt capable of denaturing the hemoglobin contained
therein and an oxidant capable of converting the
25 hemoglobin in the treated blood sample to the
met-hemoglobin form;
MSE #1838.1
X118752
b) assaying the treated sample for met-hemoglobin
as representing the amount of total hemoglobin in
the sample;
c) assaying the denatured, oxidized blood sample
5 by immunoassay for the amount of denatured form of
the particular hemoglobin derivative being sought;
and
d) calculating the relative amount of hemoglobin
10 that is in the form of the hemoglobin derivative
being sought compared to the total amount of
hemoglobin in the test sample.
This glycated protein immunoassay was found to
have a latent defect which affected its accuracy, which
15 defect is based on the discovery that the response of a
given concentration of glycated hemoglobin is not only
dependent on its concentration but is also dependent on
the total hemoglobin concentration. It has been
discovered that there is an inverse relationship
20 between the apparent glycated hemoglobin concentration
and the total hemoglobin concentration, i.e. when the
total hemoglobin concentration is high, the apparent
glycated hemoglobin concentration is less than expected
and when total hemoglobin concentration is low the
25 apparent glycated hemoglobin concentration is higher
than expected. The same phenomena is observed in the
course of determining the concentration of other
hemoglobin adducts by equilibrium analytical methods in
which the equilibrium of the dimeric form and tetrameric
30 form of the hemoglobin molecule is not perturbed by the
MSE #1838.1
X118752
- 4 -
chemistry. Immuno, enzymatic and chemical analytical
methods are regarded as equilibrium methods whereas
chromatographic and pepsin digestion methods are not.
The present invention is predicated on the discovery of
5 the underlying reason for this effect and the develop-
ment of a method to overcome it.
Summary of the Invention
The present invention is an improved method for
determination of hemoglobin adducts, e.g. hemoglobin-
10 acetaldehyde; hemoglobin-urea; hemoglobin-aspirin;
hemoglobin-glucose-6-phosphate; hemoglobin-glucose-1,6-
diphosphate; hemoglobin-glutathion and hemoglobin-
glycation end products, in a blood sample which method
comprises the steps of:
15 a) assaying the blood sample for the total
amount of hemoglobin present therein;
b) assaying the blood sample for the amount of
hemoglobin adduct present therein;
c) normalizing the measurement from the hemo-
globin adduct assay to the amount of total hemo-
globin in the sample; and
d) dividing the normalized hemoglobin adduct
concentration by the total hemoglobin concentra-
tion to obtain the corrected concentration of
25 hemoglobin adduct.
MSE #1838.1
2'118752
- 5 -
Brief Description of the Drawings
Fig. 1 is a standard curve of an immunoassay for
hemoglobin Alc. The curve was generated using six
calibrators having different percentages of HbAlc but
5 with the same hemoglobin concentration of 14 g/dL.
Fig. 2 illustrates the dependence of the observed
percentage of HbAlc in terms of total hemoglobin
concentration as determined by the immunoassay. It
also shows the second order polynomial data fitting of
10 the observed percentage of HbAlc at different HbAlc
levels.
Fig. 3 shows the al coefficients at different
levels of HbAlc. It also illustrates the linear data
fit of the a2 coefficients. The intercept of the
15 linear fit is the I2 coefficient and the slope of the
linear fit is the S2 coefficient.
Fig. 4 shows the az coefficients at different
levels of HbAlc. It also shows the linear data fit of
the a2 coefficients. The intercept of the linear fit
is the I~ coefficient and the slope of the linear fit
is the SZ coefficient.
Description of the Invention
Although the normalization procedure can be
applied to any assay which, unlike chromatography, does
25 not completely dissociate the hemoglobin molecule, the
present invention is particularly useful for the
MSE #1838.1
~1 1875 2
- 6 -
determination of glycated hemoglobin, e.g. HbAlc, by
immunoassay techniques. Other pre-equilibrium assay
methods which may be used include enzymatic, e.g. those
involving the use of oxidases, reductases or phospha-
tries that can interact with the glucose, glucose-6-
phosphate or glucose-1,6-diphosphate in the adduct; and
chemical, e.g. those involving chemical reactions that
can oxidize, reduce or hydrolyze the adduct. More
specifically, the glucose portion of the hemoglobin-
glucose adduct can be liberated as 5-hydroxymethylfur-
fural by acid treatment. The 5-hydroxymethylfurfural
is then reacted with thiobarbituric acid to form a
colored complex. The color intensity of the complex is
proportional to the concentration of the hemoglobin-
glucose adduct present in the sample. Another example
of chemical determination is the analysis of the adduct
formed by acetaldehyde and hemoglobin. Acetaldehyde,
the first metabolite of ethanol, has been shown to form
adducts with hemoglobin and the measurement of hemo-
globin associated acetaldehyde has been reported to
distinguish between drinking and non-drinking individ-
uals. The acetaldehyde in the adduct can be hydrolyzed
and then reacted with 1,3-cyclohexandeione to form a
fluorophore which is analyzed quantitatively. The
fluorophore's concentration is proportional to the
concentration of the hemoglobin-acetaldehyde adduct
present in the sample. These methods for the determi-
nation of the hemoglobin adduct typically do not
require denaturation of the hemoglobin.
The present invention is based upon the fact that
native hemoglobin exists as an equilibrium of tetrameric
MSE #1838.1
. a ..t
X11875 ~
a and (3 chains ( a2 (32 ) and the dimeric form ( a(3 ) .
These a and (3 chains are each about 143 amino acids in
length. At the end of the (3 subunit there is a valine
unit which is part of the hemoglobin Alc epitope and
can react with glucose. In a mixture of native hemo-
globin and hemoglobin adducts (as in blood) there
exists, in the case of glycated hemoglobin, a mixture
of un-glycated tetramers, un-glycated dimers, monogly-
cated tetramers and mono-glycated dimers. All of the
glycated components will have different immuno, enzymatic
or chemical reactivity; the glycated dimers have the
highest reactivity since their glycation sites are
exposed and accessible to interaction with the reactant.
The glycated tetramers have much lower reactivity
because in the tetrameric configuration the glycated
site is sterically hindered. When the total concen-
tration of hemoglobin in the sample increases, the
equilibrium between the tetrameric form of glycated
hemoglobin and its dimeric form is shifted towards the
tetrameric form because at higher concentrations, the
associated, tetrameric form is thermodynamically
favored. The combination of this equilibrium shift and
the differences in reactivity between the tetrameric
and dimeric forms of glycated hemoglobin with the
antibody reagent cause the assay to be affected by the
total amount of hemoglobin present in the blood sample.
In the clinical setting, this "hemoglobin dependency
effect", which is a bias based on the total amount of
hemoglobin present in the sample, should be taken into
account when the concentration of the hemoglobin adduct
is being determined.
MSE #1838.1
~''~ 18 7 5 ~
-8_
Depending on the hemoglobin adduct whose presence
or concentration is being sought and the particular
method used in its assay, it may be necessary to expose
the reactive site of the hemoglobin adduct by denatur-
5 ation. The term denaturation, as used herein, is
intended to mean any treatment which will expose the
reactive site to the analytical reagent selected for
use without disrupting the equilibrium between the
dimeric and tetrameric hemoglobin chains. Suitable
10 denaturation techniques include treating the sample
with a high salt concentration; low or high pH; a
choatropic reagent such as thiocyanate, guanidinium,
urea or a surfactant such as a non-ionic, cationic or
anionic detergent. All of these treatments cause the
15 disruption to some degree of the hydrogen bonding of
the hemoglobin molecule thereby exposing the reactive
site. These treatments do not cause the complete
dissociation of the hemoglobin and, therefore, do not
alter the aforementioned equilibrium. Proteolytic
20 digestion or chromatography would not be suitable for
use in conjunction with the present method since it
will disrupt this equilibrium. The determination of
glycated hemoglobin, e.g. HbAlc, will typically involve
denaturation of the hemoglobin, especially when an
25 immunoassay technique is to be employed.
The method of the present invention, in which an
immunoassay and a glycated hemoglobin are chosen for
purposes of illustration, is carried out as follows:
In the first step, the hemoglobin is denatured in
30 order that the glycated N-terminal peptide residue
MSE #1838.1
X118752
_ g _
becomes available for antibody binding. In United
States Patent 4,658,022 there are described numerous
techniques for protein denaturation including treatment
of the protein by physical means such as heat, sonica-
tion, high and low pH and chemical denaturation by di-
gestion or interaction with a chaotropic agent such as
guanidine, urea or a detergent such as sodium dodecyl-
sulfate. In a preferred technique, the protein is de-
natured to expose the peptide epitope for antibody bond-
ing by treating the blood sample with a thiocyanate salt
and an oxidant as disclosed in U.S. Patent 4,970,171.
The technique taught in this patent oxidizes the hemo-
globin to met-hemoglobin which is converted to thiocyan-
met-hemoglobin as the stable color component, i.e. the
color intensity is correlated to the concentration of
the hemoglobin. The resulting thiocyan-met-hemoglobin
serves as the basis for measuring total sample hemo-
globin with the denatured form of the particular
hemoglobin derivative serving as the analyte in the
immunoassay part of the procedure. The thiocyanate
salt is selected from those salts which upon ionization
provide the thiocyanate anion (SCN-) to render the
hemoglobin suitable for detection in the immunoassay.
Ammonium, sodium and potassium are suitable counter-
cations. Lithium thiocyanate provides faster denatur-
ation as well as expedited lysing of red blood cells in
the blood sample being tested. The oxidant, which can
be essentially any inorganic oxidizing agent, converts
the native hemoglobin ferrous ion to its ferric met-
hemoglobin form. Suitable oxidants include ferricyanide,
a
:~~ 185 '~
- 10 -
iodate, chlorate, bromate, chromate, hypochlorite,
iodate, periodate and peroxide.
Met-hemoglobin concentration is conveniently
determined by measuring its characteristic absorbance
at 540 nanometers using a conventional spectrophotometer
such as the HP 8450A spectrophotometer or its equivalent.
Alternatively, one can use the Drabkin's procedure as
described in National Commission for Clinical Laboratory
Standards of the United States, Proposed Standard
PSH-15.
After denaturation, the blood sample is assayed
for total met-hemoglobin and for the hemoglobin adduct
of interest. This is most conveniently accomplished by
immunoassay using monoclonal antibodies developed to
bind specifically to the epitope that characterizes the
hemoglobin adduct, e.g. hemoglobin Alc, in the denatured
protein. The particular immunoassay technique and
format, as well as the labeling approach and detection
signal generated, is not critical. Radioisotopic or
enzyme labeled (ELISA) techniques can be used. A
particle agglutination inhibition immunoassay based on
the specific interaction of an antibody particle
reagent and an agglutinator reagent is particularly
useful. The antibody particle reagent comprises the
monoclonal antibody, or a fragment thereof, bound to a
water suspensible particle, typically a polystyrene or
other latex. The agglutinator comprises a plurality of
epitopic binding sites for the antibody reagent. In
the absence of analyte, the antibody particle and
agglutinator bind to one another to form a light
MSE #1838.1
- 11 - ~~ 1875 ~
scattering complex that can be quantified by turbidi-
metric measurement. In the presence of increasing
amounts of analyte, the turbidity of the solution
decreases as antibody particles become bound to the
5 analyte and cannot bind to the agglutinator which is
typically a polymer backbone to which is attached a
number of organic moieties, such as peptide residues,
which define the epitope that characterizes the hemo-
globin derivative of interest. When hemoglobin Alc is
10 to be determined, the epitope comprises the glycated
peptide residues of a few amino acid units corresponding
to the sequence of the glycated N-terminal residue in
hemoglobin Alc.
After the concentration of total hemoglobin and
15 the hemoglobin derivative are determined, the percentage
of the hemoglobin adduct as a portion of total hemo-
globin is calculated. This determination, as taught in
the prior art, is carried out by simply dividing the
hemoglobin derivative concentration by the concentration
20 of total hemoglobin to arrive at the percentage value
of the hemoglobin adduct. This calculation does not
provide a totally accurate picture, however, due to the
previously mentioned hemoglobin dependency effect. It
is an object of the present invention to provide a
25 Procedure for factoring the hemoglobin dependency
effect out of the determination of the hemoglobin
derivative to thereby provide a more accurate procedure
for determining the concentration of the hemoglobin
derivative under consideration. This improved procedure
30 involves normalizing the concentration of the hemoglobin
derivative as determined by the assay. This is
MSE #1838.1
12 _ X11875 2
accomplished by deducing the true concentration of
hemoglobin derivative from its reactivity and the total
concentration of hemoglobin, both of which values are
determined during the assay. The normalization procedure
5. is carried out in 5 steps. These steps are:
A. Creating a calibration curve relating (a) the
concentration of the hemoglobin adduct as determined by
a reference procedure, such as HPLC, which is not
subject to the hemoglobin effect to (b) the immuno
10 reaction response by using calibrator blood samples
containing a normalized concentration of hemoglobin and
different concentrations of the hemoglobin adduct;
B. Preparing a series of blood samples containing
different concentrations of hemoglobin but the same
15 percentage of hemoglobin adduct in each series, deter-
mining the immunoreaction response from each sample and
obtaining the observed concentration of hemoglobin
adduct for each sample from the calibration curve
generated in Step A;
20 C. Creating a series of polynomial (n°'' order)
fitted curves relating (a) the observed concentration
of hemoglobin adduct from each series as determined in
Step B to (b) the concentration of hemoglobin in each
sample in the series, obtaining the polynomial curve
25 fitting coefficients (an, an_1, an_2....ao) for each
curve and repeating the process for the whole series of
curves;
MSE #1838.1
'~ 18 7 5 ~
- 13 -
D. Obtaining the linear curve fitting relating
(a) the polynomial curve fitting coefficient an for the
series of curves generated in Step C to (b) the percent-
age of hemoglobin adduct at the normalized concentration
5 of hemoglobin and obtaining the slope (Sn) and intercept
(In) of the linear curve fit;
E. Obtaining the Sn_1, Sn_2 ....51 and In_1,
In_2 .... I1 values from the polynomial curve fitting
coefficient an_1, an_2 .... al respectively by the
10 procedure of step D; and
F. Obtaining the corrected hemoglobin adduct
concentration by relating (a) the observed hemoglobin
adduct concentration to (b) hemoglobin concentration by
solving the equation:
15 Hbad = (Hbad' - (Hbr'-14n)*Ir,-(Hbn-1-14"-1)*Ii,_,_
(Hbn-2-14"-2)*I"_2-.......(Hb-14)*I1)~((Hbn-14n)-
*Sn+(Hbn-1-14"-1)*Sn_1+(Hbn-~-14n-2)*Sn_2+.......-
(Hb-14)*Sl+1).
where:
20 Hbad is the corrected concentration of the hemo
globin adduct whose concentration is being sought;
HbHd' is the observed concentration of hemoglobin
adduct as determined by immunoassay;
Hb is the total hemoglobin concentration in terms
25 of g/dL;
MSE #1838.1
.~. ~ '~ 18 7 5
- 14 -
In...Il and Sn...Sl are experimentally determined
coefficients; and
where 14 represents 14 g/dL hemoglobin selected as
the normalized concentration. Normalized concentration,
as used herein means a fixed concentration of hemoglobin.
Any concentration can be selected as the normalized
concentration, however, 14 g/dL is preferred for use in
this procedure because the majority of clinical blood
specimens will contain this concentration of hemoglobin.
The advantage in this is that, if the specimen's
hemoglobin concentration is equal to the normalized
concentration, no correction is required, since, in
that case, the corrected concentration is equal to the
observed concentration.
The method of practicing the present invention is
further illustrated by the following example in which
HbAlc is the hemoglobin adduct of interest. In this
example the second order polynomial is selected for
illustration purposes. One can use a linear, 2"d order
2p or higher order polynomial curve fit in this procedure.
However, the linear curve fit will not correct the
curvature of the results and the algorithm for higher
order polynomials will be more complicated and carry
more coefficients, i.e. the linear fit will require
coefficients I1 and S1; the second order polynomial
will require coefficients Il, I2, Sl and S2 while the
third order polynomial will require coefficients I1,
I2~ Is. S~. SZ and S3.
MSE #1838.1
_ 15 - X11875 ~
EXAMPLE
A. Six calibrators were prepared containing
2.5%, 5.9%, 7.9%, 8.93%, 10.97% and 12.54% hemoglobin
Alc respectively as determined by HPLC. The total
5 hemoglobin concentration in each calibrator blood
sample was 14 g/dL. The immunoreaction response of
each calibrator was measured by the latex agglutination
technique previously described with the results being
set out in Table I.
10 TABLE I
%Alc 14 g/dL
Alc (mM) mA
0 0 827.9
2.5% 108.5 670.5
15 5.93% 257.4 440.2
7.97% 345.9 321.7
8.93% 387.6 271.6
10.97% 476.1 212.3
12.54% 544.2 173.4
20 The immunoreaction responses were used to generate
the calibration curve of Fig. 1 with the immunoreaction
responses being used as the Y values and the reference
Alc concentrations used as the X values. In this
example, the immunoreaction response is represented by
25 mA (540 nm, 168-28) which is the difference in milli-
absorbance reading at a wavelength of 540 nanometers
between 28 and 168 seconds after start of the reaction.
B. The next step was to prepare six groups of
samples containing 2.5%, 5.93%, 7.97%, 8.93%, 10.97%
MSE #1838.1
- 16 -
and 12.54 Alc (corresponding to the HbAlc levels found
by the reference method in step A) each of which group
having 5 or more levels of hemoglobin concentration
between 9 and 20 g/dL. These samples were prepared
from packed human red blood cells and human plasma.
Their immunoreaction responses, the Alc concentrations
of these samples were calculated using the calibration
curve generated in Step A. The sample concentrations
are shown in Table II as observed Alc values. The
procedure for measuring HbAlc is described in detail in
U.S. Patent 4,847,209 as well as how wM HbAlc relates
to % HbAlc.
C. For each group of samples in Step B, the
observed ~ Alc values were plotted against the hemo-
globin concentration (Fig. 2). Using the observed ~
Alc values as Y values and the hemoglobin concentration
as X values (Table II), the data can be fitted with the
polynomial curve fit using a commercial curve fitting
program such as SlideWrite~ from Advance Graphic
Software Inc. or Sigmaplot~ from Jandel Scientific.
The data can be fitted with any order polynomial.
For the group of 2.5o samples, the observed % Alc
values of 2.28, 2.43, 2.57, 2.43, 2.75 and 2.5 were
used as the Y values and the hemoglobin concentration
of 21.9, 17.5, 14, 14, 12.1, 10.4 were used as the X
values. After the 2nd order polynomial curve fitting
of these data, the equation Y=-0.0011617*XZ+0.02542*-
X+2.490 was generated for the group of samples.
k.-,
a,
~11875~
- 17 -
For the group of 5.93% samples, the observed % Alc
values of 5.17, 5.58, 5.79, 6.06, 6.35 and 6.45 were
used as the Y values and the hemoglobin concentration
of 19.9, 16.3, 14, 13.2, 11.7 and 10.3 were used as the
5 X values. After the 2~d order polynomial curve fitting
of these data, the equation Y=0.004478*Xz-0.2747*-
X+8.859 was generated for this group of samples.
For the group of 7.97% samples, the observed % Alc
values of 7.05, 7.88, 8.05, 8.26, 8.87 and 9.08 were
10 used as the Y values and the hemoglobin concentration
of 20.4, 16.8, 14, 13.4 11.6 and 10.4 were used as the
X values. After the 2nd order polynomial curve fitting
of these data, the equation Y=0.006158*X2-0.3860*-
X+12.419 was generated for this group of samples.
15 For the group of 8.93% samples, the observed Alc
values of 7.92, 8.60, 9.33, 9.26, 9.92 and 10.58 were
used as the Y values and the hemoglobin concentration
of 21.9, 17.1, 14.2, 14, 12.3 and 10.8 were used as the
X values. After the 2nd order polynomial curve fitting
2p of these data, the equation Y=0.017231*X2-0.7985*-
X+17.158 was generated for this group of samples.
For the group of 10.97% samples, the observed %
Alc values of 9.72, 10.94, 11.02, 12.14, 12.96 and
14.13 were used as the Y values and the hemoglobin
25 concentration of 19.7, 15.2, 14, 12, 10.4, 8.7 were
used as the X values. After the 2"d order polynomial
curve fitting of these data, the equation Y=0.01644*-
X2-1.14462*X+11.064 was generated for this group of
samples.
MSE #1838.1
~11875~
- 18 -
For the group of 12.54% samples, the observed %
Alc values of 10.13, 11.97, 12.44, 13.08, 14.46 and
16.01 were used as the Y values and the hemoglobin
concentration of 20.9, 16.2, 14, 13, 11 and 9.2 were
5 used as the X values. After the 2nd order polynomial
curve fitting of these data, the equation Y=0.02734*-
X2-1.3086*X+25.608 was generated for this group of
samples.
In these second order equations; -0.001617,
10 0.004478, 0.006158, 0.017231, 0.02644 and 0.02734 are
the a2 coefficients and 0.02542, -0.2747, -0.386,
-0.7985, -1.1462 and -1.3086 are the ai coefficients
generated from the curve fitting steps. The procedures
of steps D and E require the use of these al and as
15 coefficients for further calculation.
D. The al values, 0.02542, -0.2747, -0.3860,
-0.7985, -1.1462 and -1.3086 were used as the Y values
and plotted against the % Alc values of 2.5, 5.93,
7.97, 8.93, 10.97 and 12.54 as the X values to prepare
20 Fig. 3. The data were fit with a linear curve fit
using the Slidewrite~ curve fitting program. This
linear curve fitting gives the result of:
Y = -0.141*X + 0.505
where, -0.141 is the Sl coefficient and 0.505 is the Il
25 coefficient.
E. The a2 values -0.001617, 0.004478, 0.006158,
0.017231, 0.02644 and 0.02734 were used as the Y values
MSE #1838.1
~118~5~
- 19 -
and plotted against the % Alc values of 2.5, 5.93,
7.97, 8.93, 10.97 and 12.54 as the X values to prepare
Fig. 4. The data were fit with a linear curve fit to
give the result of:
Y = 0.0032*X -0.0129
where 0.0032 is the S2 coefficient and 00.0129 is the
I2 coefficient.
F. The corrected Alc concentration was obtained
by solving the equation:
Alc = (Alc'-(Hb2-14~)*I~-{Hb-14)*I1)/((Hb2-14~)-
*Sa+(Hb-14)*S1+1)
(Equation I)
where:
Alc' is the observed % HbAlc as determined by the
i~unoassay;
Hb is the total hemoglobin concentration in terms
of g/dL;
I2, Il, SZ and Sl are experimentally determined
coefficients, and
14 is 14 g/dL hemoglobin selected as the normalized
concentration.
MSE #1838.1
_ 20 _ ~ 1187 5 ~
When the concentration of hemoglobin is further
away from the normalized concentration, more bias of
the HbAlc value from the reference value (e. g. HPLC
value) is observed. Therefore, it is advantageous to
5 set the normalized concentration to the concentration
of hemoglobin that is most commonly encountered in the
blood sample being examined to provide a test having
the least bias. This correction method can be used
with various normalized concentrations, e.g. 13 g/dL,
10 14 g/dL, 15 g/dL, etc. However, since the majority of
blood specimens will have a hemoglobin concentration of
14 g/dL, it is advantageous to select this level as the
normalized concentration.
In the foregoing example there is described the
15 determination of the hemoglobin Alc epitope, however,
the technique of the present invention is equally
applicable to the determination of other hemoglobin
derivatives, such as those having other glycated amine
epitopes along the a and (3 subunits. Likewise, the
20 present method is applicable to assay methods other
than immunoassay which do not completely dissociate the
hemoglobin. These methods include chemical and enzymatic
assays.
Referring to Table II which reports the results of
25 this experiment, the bias (% difference between the
observed value and the reference value) of the uncorr-
ected results ranged from 28.7% to -18.6%. After
correction with the algorithm (Equation I) and the
above coefficients, the bias were reduced to 7.0 to
30 -5~6%.
MSE #1838.1
X11875 ~
- 21 -
c
~o d o
O U ~.i
~ C w
~ 07 U
O H N
~ w N v
C N O
O fx U
E 4
ro O d
C S.I a~
~.-1 w w ~ tp O ~ O O O N O V~ !~~ N .-a OD O ~ .-~ .H c~7 ~p t(1 O ~f7 W '~1
~ O If1 O t0 t~ ~ O O~ O 1(1
. . . . . . . . . . . . . . . . . .
a 0J d' u1 O 1l1 t~ .~ N N O N N O c~1 ~O O O .~ .r M N .H O O O a' t~'1 O .w
.-~ .H .H ~. O O N ~'1
~roro m
~, a~
as a
ev ro
w
0
~~~~ H C
a.~ O O
~ 1~~ ~i O~ V' N O~ N O ~'1 p V' Il1 01 !~
~ W Y~ ~D N1 lW '1 If1 O O N O~ M ~l1 O O O 1(1 01 VW p tp O O ~O O t'f V~ ~ O
.H .H N N Ov V~ If1 v0 m
V'ep~!'laTt~~O 0~01P0~01~ t'~1t000.-101 ~t1111V'Nt'~1~'1 ~.~'.i.r.-r.-1 NNNNNN
C U H N N N N N N If1 V1 tf1 1l1 It1 If1 W W ~ ~ ~ n 01 01 01 01 01 01 .-I .-~
.H .-r .-1 .-i .,1 .-1 .-1 .-1 .m-1
a ..~ H
o~ o
U
do
x
m a~
~ d
ro a
a E .~
oro
H ~ I~ ~D O t~ f~ er V' .-~ O t0 N m 1f7 .-~ W O .-1 f~7 Q) n O N t0 N 10 W O
N N !~
~ W M at O V~ N P
d ' ~ ~ ~ O f~'7 O V~ Of .-i N N O N O N ~ ~ O O !~ ~ H O O O t~ 0i ~u tai Ci
~t5 ip ~
C ,.wl ~ O lit v0 N .r .r .r .H .-I .-~ .r .H .~ .-r N .-I .-~ N
I I I I I I I I I I
C. M 4) I I I I
w f~ H
8 41
Hsnw
m Q1
~ 0;
m
O
U sn vi
.r N
wva W
.O C~ ..~ m M n r~i r~ ~ 00 0»o um nn oo u~wo n oo N r»o N ao N ~ N wo M M n
v' ao ~o .-~ p
x H~ Nmn~mn .,me~och~r oaooNaoo owoehNOmn ~a~o',o,.., .~o~a~oa~o.i
H C) N N N N N N if1 1l1 ~(1 ~p tp ~p ~ t~ p 07 0) O~ t' 0) 01 0~ 0~ O 0~ O
..r N N V~ O ~ N f~1 at t0 ,7
N ..i N .~ ..1 r1 .~ .H .~ r.1 .~~
o~ a
C 'v
o w
.. o,
U .c C
a~ x o
H
Hb~
o a~ ro on
OW f1 .-r d' OW '~1 N I~ t~7 V~ CD d' ~D V' 0~ H N ~'~1 CD l~ N ~ n 0~ N N ~.i
C d C .r !~ V~ ~ N O 01 t0 e! f~'1 .-~ O O tp V~ f~'7 ri O .-a I~ et ~ N O 0~
V1 V~ N O 0) O t0 V~ ~'1 .d 01
.r tn d N .H .-v .-~ .~ .-~ .-v .r .r ..~ .H .w
17 17 U N .~ .r .~ .-~ .~ N .~ .-i .~ .~.~ .r .-.~ ..~ .r .-~ ~ N ..1 .-I .-i
.r
O O C
~~ O 'O
U GI
O U a~
E ~~ ro
O ~ .i
x a
0l U
U dp dp ~ .-i
C dP fn ~ ~ ~ Id
H N ~ ~ °; ~ °~ ~ U
w ,~ '~' '~ c~ oo O cv
a~ ro ~ ~ -a
x > ca
MSE #1838.1
--
X11875
- 22 -
The normalization for hemoglobin concentration
step of the present invention can be carried out by
means of a pre-programmed computer memory which may be
in the form of a microchip. Incorporating the microchip
into a known device for the determination of a prese-
lected hemoglobin adduct, such as that described in
Diabetes Care, D.G. Marrero, et al 15/8 1045-1049
(1992), provides a device comprising
a) means for determining the total amount of
hemoglobin in the blood sample;
b) means for carrying out an assay of the
hemoglobin adduct;
c) means for normalizing the measurement of the
hemoglobin adduct to the total amount of hemoglobin
in the sample; and
d) means for dividing the normalized hemoglobin
adduct concentration by the total hemoglobin
concentration to obtain the corrected concentra-
tion of the hemoglobin adduct.
2Q When the hemoglobin adduct involved and/or the
assay method are such that denaturation of the hemoglobin
are required, the device will also include means for
denaturing the hemoglobin.
The instrument contains a memory which is prepro-
grammed to do all of the calculations necessary for
normalizing the measured concentration of the hemoglobin
MSE #1838.1
_ ~11875~
- 23 -
adduct to the total amount of hemoglobin in the blood
sample.
The mathematical equation, e.g. Equation I and the
I and S coefficients can be programmed into and stored
in the memory of a calculation device such as a calcu-
lator, computer or computer that has been interfaced
with a clinical analyzer capable of determining the
total concentration of hemoglobin and that of the
hemoglobin adduct. For each clinical sample, the
observed concentration of hemoglobin adduct and total
hemoglobin concentration is determined by running the
assay. The corrected hemoglobin adduct concentration
is then determined either manually using the calculator
or by means of the pre-programmed computer. Alterna-
tively the results can be obtained directly from a
computer that has been interfaced with the clinical
analyzer.
MSE #1838.1
