| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Division of Clinical Chemistry, Department of Pathology, and
2
Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University School of Medicine, 600 N. Wolfe St., Baltimore, MD 21287-7065.
a Address correspondence to this author at: Dallas Veterans Affairs Medical Center, 4500 Lancaster Road, 113, Dallas, TX 75216. Fax 214-857-0739; e-mail Martin.Kroll{at}med.va.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Methods: Interference of a polymerized hemoglobin blood substitute with the assay of alkaline phosphatase was examined with a Hitachi 917 analyzer and ultraviolet-visible spectrophotometry.
Results: Hemoglobin-based oxygen carrier solutions had substantial absorbance at 415 nm, the wavelength of analysis used to measure the formation of 4-nitrophenol. In addition to offsetting the initial absorbance at the analytical wavelength, polymerized hemoglobin gave rise to a strong negative interference plot because of alkali denaturation of the substitute. The same interference mechanism was also observed for native hemoglobin (hemolysate), indicating that the interference was not derived from the polymerization process. The interference can be corrected by implementing a rate-correction procedure, or the interference can be avoided by measurement at 450 nm.
Conclusions: The interference of polymerized hemoglobin in the alkaline phosphatase assay is a result of an absorbance offset caused by alkali denaturation of hemoglobin. The interference can be corrected or avoided by modifying the calculation or the analytical wavelength. The correction strategy may also be applicable to improving the hemolysis index for this method.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Although there are reports on interferences caused by HBOC (4)(5)(6), little is known about the mechanisms of interference. It may be simplistically assumed that the interferences are purely spectral in nature because of the broad absorbance spectrum of polymerized hemoglobin. However, in internal studies we have noted a wide variability in interference plots seen with different analytes and different analyzers, which indicates that it is likely other factors are involved. Understanding the mechanism of HBOC interference will assist both manufacturers and laboratorians, and offer possibilities to correct the interference.
We performed a variety of spectroscopic studies to elucidate the mechanism of interference caused by HBOCs in a commonly used enzymatic test method for alkaline phosphatase. The studies confirmed that the broad background absorbance spectrum imposed by polymerized hemoglobin is not the only factor involved in the generation of interference. We present two strategies to render the method interference free, based on an understanding of the mechanism involved. These studies may also serve as a model for the investigation and correction of blood-substitute interferences in other clinical chemistry test methods.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
0.93 mol/L
2-amino-2-methyl-1-propanol (AMP), pH 10.5, and 1.04 mmol/L
magnesium-L-aspartate as reactive ingredients, and
hydrochloric acid and zinc sulfate heptahydrate as nonreactive
ingredients. The R2 buffer/substrate reagent also contains 97.9
mmol/L 4-nitrophenylphosphate. A commercially developed polymerized hemoglobin solution was used for all interference studies (PolyHeme; Northfield Laboratories). This blood substitute consists of native tetrameric human hemoglobin polymerized using glutaraldehyde and modified by pyridoxal phosphate to optimize the P50 value. The commercial solution has a hemoglobin concentration of 100 g/L and is essentially free of all unreacted tetramer. The latter claim was verified by gel filtration chromatography fractionation using a 2.5 x 50 cm column packed with Sephadex G-200 gel (Sigma) and elution with 0.1 mol/L phosphate buffer, pH 7.4. The molecular weight distribution of the materials was determined using a molecular weight marker kit (MW-GF-1000 kit; Sigma) with protein markers spanning the range Mr 29 000–2 000 000. Native hemoglobin was obtained in-house from a standard hemolysate. Refrigerated serum specimens were used as the source of alkaline phosphatase.
Samples for the interference studies were prepared by adding blood substitute and saline to serum. Three series of solutions were prepared with different alkaline phosphatase activity (low, 60 U/L; normal, 100 U/L; high, 160 U/L). The HBOC concentration in each series was 0–50 g/L. The interference samples were analyzed for alkaline phosphatase on the Hitachi 917 in standard mode. Test result and absorbance vs time data were accessed via the software module of the analyzer. Spectrophotometric studies were also performed using a Lambda 3B ultraviolet (UV)-visible spectrophotometer (Perkin-Elmer) with selected interference study samples, in either fixed wavelength or scan mode. Data from all studies were imported into a statistical software program for further analysis.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
 |
As a first step, we examined the molecular weight distribution and
UV-visible absorbance spectrum of polymerized hemoglobin. Absorbance
measurements (280 nm and 415 nm) of the fractions obtained from
gel-filtration chromatography demonstrated that the blood substitute
contains a distribution of hemoglobin polymers with molecular weights
of 100 000–600 000. No detectable absorbance was found in the
fractions where native hemoglobin elutes
(Mr 
65 000). The glutaraldehyde
cross-linking produces a spectrum that is still essentially identical
to native hemoglobin, with oxyhemoglobin peaks in the 540–580 nm
region and a large Soret band near 410 nm (Fig. 1
). This latter band, however, overlaps the 4-nitrophenol peak
absorbance at the analytical wavelength of 415 nm.
|
It might be expected that because the alkaline phosphatase assay is a
kinetic rate method, the presence of HBOC would offset the absorbance
readings but not necessarily influence the rate of product formation
measured at 415 nm. However, interference plots for alkaline
phosphatase (Fig. 2
) show that as the concentration of HBOC in the serum specimen
increases, there is a linear decrease in the Hitachi 917 test result.
This observation is indicative of an additional mechanism that
attenuates the rate of increase in the measured absorbance at 415 nm. A
test result less than the limit of detection is obtained when there is
no alkaline phosphatase activity present, regardless of HBOC
concentration. The interference plot is also consistent with the claim
that the Hitachi 917 alkaline phosphatase assay exhibits significant
negative interference from hemoglobin at a hemolysis index (H index)
>2 g/L (200 mg/dL) (7).
|
The test result data from the interference plots (Fig. 2
) were
evaluated using a statistical software package (Statmost, Ver.2.5;
DataMost Corp.). Multiple regression analysis failed to find a
significant coefficient for the cross-term (8), indicating
that there is no interaction between enzyme and blood substitute. The
parallel interference plots in Fig. 2
also support this conclusion,
demonstrating that the magnitude of the interference is a function only
of the HBOC concentration and is independent of the activity of
alkaline phosphatase. The interference from the blood substitute is
approximately constant at -50 U/L alkaline phosphatase per 10 g/L
HBOC, regardless of the activity of alkaline phosphatase. Furthermore,
Lineweaver-Burk plots (not shown) constructed from studies performed on
the spectrophotometer did not indicate either a purely competitive or
noncompetitive enzyme inhibition model. Taken together, these
observations are consistent with a mechanism in which the blood
substitute acts as an independent contributor of interference in this
method.
We sought to further elucidate the mechanism by examining the
absorbance data accessible through the software module of the Hitachi
917. The alkaline phosphatase assay consists of a total of 34
measurement points, each spaced by 20-s intervals. The analyzer
software multiplies the absorbance scale by a factor of 10 000 for
greater resolution (see Table 1
and Fig. 3
). For an ordinary sample, the incubation interval (first 16
measurement points) of sample with buffer (R1) is relatively flat with
absorbance values close to zero. Addition of substrate solution (R2) at
point 17 shifts the reaction plot up because of absorbance of substrate
(in excess) at 415 nm. A positive slope is observed thereafter because
of enzymatic formation of product, and the rate of change of absorbance
measured between points 25 and 34 is used to calculate the enzyme
activity. For a sample containing HBOC, the absorbance readings are
offset because of the strong absorbance of the blood substitute at 415
nm, and a negative slope is observed during the incubation interval.
After addition of substrate at point 17, the sample with blood
substitute exhibits an apparent decreased rate of product formation, as
indicated by a lower slope during the measurement interval.
|
|
As mentioned earlier, absorbance values used for activity calculations by the Hitachi 917 software are the difference between the primary wavelength absorbance reading (415 nm) and the secondary wavelength absorbance reading (660 nm). As observed at 415 nm, the absorbance also increases at 660 nm in proportion to the amount of HBOC present. However, these increases are negligible (<1%) compared with the increases observed at 415 nm. Furthermore, it was demonstrated (see later in text) that there is zero drift at 660 nm in the presence of HBOC. We have used the difference values in our calculations and plots, although the results are identical when only the primary wavelength is used. The interference is independent of changes occurring at the secondary wavelength.
The effect of blood substitute on the Hitachi 917 measurement point
data was examined in closer detail using a serum specimen with an
alkaline phosphatase activity of 500 U/L. The decrease in the apparent
rate of product formation observed during the measurement interval
(Fig. 3A
, points 25–34) is roughly a constant function of the HBOC
concentration, consistent with previous observations. The decline in
absorbance values observed during the incubation interval (Fig. 3B
, points 1–16) is also a relatively constant function of the HBOC
concentration. The absorbance values for the sample with no blood
substitute are relatively stable, whereas increasing amounts of HBOC
produce an increasingly negative slope during this interval. The same
effect is observed both with and without the presence of serum/alkaline
phosphatase. In particular, the negative slope does not reach a plateau
at the end of the incubation interval, but continues into the
subsequent measurement interval after addition of substrate. This was
readily observed in specimens containing blood substitute but no
alkaline phosphatase activity.
We examined the possibility that the decreasing absorbance of the
polymerized hemoglobin in the buffer reagent is attributable to a pH
effect. To test this hypothesis, studies were performed with a
different buffer than the Roche AMP buffer reagent. A constant amount
of HBOC was added to a series of phosphate buffer solutions spanning
the pH range 4–11. After addition, the absorbance at 415 nm was
measured with the spectrophotometer for a 10-min interval. The blood
substitute is relatively stable in the pH range 5–10 but exhibits a
marked negative shift in absorbance at both extremes of pH (pH 4 and
11). In particular, there is a rapid reorganization or denaturation
that occurs above pH 10, consistent with the negative drift in
absorbance on the Hitachi 917 analyzer, which uses an AMP buffer with a
pH of 10.5 (Fig. 4
). Of interest is our observation (in separate studies) that the
alkaline phosphatase/AMP assay used on the CX-7 (Beckman) also has a
negative interference plot with this blood substitute, but the
interference is less pronounced than that observed in the Hitachi 917
method. Because the CX-7 method is essentially identical to the Hitachi
917 method (same reagents and same wavelength) but uses an AMP buffer
with a lower pH of 10.3, it is possible that the reduced interference
on this analyzer is attributable to a reduced rate of denaturation at
the lower pH.
|
Overall, these results indicate that the interference caused by the declining absorbance of HBOC at 415 nm is attributable to a pH effect rather than a chemical interaction between blood substitute and components of the buffer solution. Furthermore, interference studies done with native hemoglobin gave results identical to those with the blood substitute, indicating that the observed interferences are mainly attributable to the hemoglobin units in the blood substitute and not to any special properties that may have been introduced as a result of the polymerization process.
To correct for the interference caused by the presence of blood
substitute, a rate-correction procedure was investigated. The Hitachi
917 alkaline phosphatase activity (C) result is calculated
by the formula C =
k
A25–34, where
k is a constant and A is the change in
absorbance per minute (rate) observed during the measurement interval
(measurement points 25–34, which represent a 3-min time period). If it
is assumed that the decrease in
A25–34 seen when HBOC is present
is because of an offset caused by the simultaneous denaturation of
hemoglobin, it becomes apparent that the negative rate observed during
the incubation interval could be applied as a correction. However, it
was observed that this rate gradually changes over time (Fig. 3B
).
Therefore, to approximate the rate of alkali denaturation during the
measurement interval, we examined the difference between the fourth
(measurement points 10–13) and fifth (points 13–16) minutes of the
incubation interval and calculated:
 |
Acorrected then returns
a corrected value for alkaline phosphatase. The results are presented
in Table 1
. Use of this algorithm significantly improves the reported
values, allowing the method to return results within 15% of the
expected value up to a HBOC concentration of 30 g/L. The rate
correction method fails at higher concentrations of blood substitute
because of poor photodiode detector performance when transmitted light
is low (excessive absorbance). No effect is expected from the
additional buffering capacity from high concentrations of hemoglobin
because the concentration of AMP buffer is always several orders of
magnitude higher.
The use of an alternate wavelength at which drift attributable to
alkali denaturation of hemoglobin might be negligible was also
considered as a possible strategy to avoid interference from the blood
substitute. To identify a suitable wavelength, HBOC was added to AMP
buffer and the absorbance spectrum was measured on the
spectrophotometer at t = 0 and t = 10
min. The difference of these two spectra is plotted in Fig. 4
. As
expected, peaks characteristic of hemoglobin underwent a negative shift
during the 10-min interval. Incidentally, we noted that drift at the
secondary wavelength used in the Hitachi method, 660 nm, was close to
zero. We found that use of the region near 320 (340) nm was prohibited
by excessively strong absorbance from the enzyme substrate. We
therefore selected 450 nm as a wavelength at which product absorbs
(Fig. 1
) but at which drift attributable to alkali denaturation should
be close to zero. The wavelength 450 nm was selected instead of 440 nm
because it is a wavelength that frequently is available in commercial
chemistry analyzers.
The performance of the alkaline phosphatase method at 415 nm vs 450 nm
is shown in Fig. 5
. The studies were performed on the spectrophotometer, using the
Hitachi 917 reagents and allowing a 5-min incubation interval of serum
in AMP buffer before addition of substrate. The activity of alkaline
phosphatase was the same in all experiments (150 U/L). After the
addition of substrate and allowing a 3-min lag phase, we found that the
subsequent 4-min change in absorbance at 415 nm showed a large decrease
with increasing HBOC concentrations, but that the absorbance at 450 nm
was relatively stable (Fig. 5A
). The 4-min change in absorbance was
chosen as the measured variable rather than the change in
absorbance per minute to provide larger numbers and to improve the
resolution in the comparison. When the data are presented in terms of
the percentage difference in the 4-min absorbance change from that
observed when no blood substitute is present (Fig. 5B
), the use of 450
nm as the analytical wavelength returns a result within 15% of the
expected value and permits the method to be used up to a HBOC
concentration of 50 g/L.
|
In this report, our intention was to identify and confirm the mechanism of interference of a polymerized hemoglobin blood substitute in an alkaline phosphatase method commonly used on clinical chemistry analyzers. The blood substitute we examined (PolyHeme) may be useful as a safe alternative to allogeneic red blood cells in transfusion medicine (9), and would seem to be poised to gain market approval after the completion of current phase III clinical trials. This blood substitute is different from its predecessor (dissociable tetrameric hemoglobin) in having minimal toxicities of renal dysfunction and vasoconstriction. However, as an interferent it behaves remarkably similarly to native hemoglobin. This similarity is evident in both the UV-visible absorbance spectrum and in the interference mechanism observed in the alkaline phosphatase assay.
The mechanism of interference of polymerized hemoglobin in the alkaline phosphatase assay involves an alkali denaturation of hemoglobin, which causes a decreasing absorbance at 415 nm, which in turn decreases the apparent rate of product formation after the addition of substrate. The interference appears to depend solely on the concentration of blood substitute and not on an interaction with reagent components, causing a constant interference of approximately -50 U/L per 10 g/L HBOC. The phenomenon of alkali denaturation has in fact been exploited previously in the measurement of fetal hemoglobin (10)(11), in which addition of sodium hydroxide is used to denature and ultimately remove all hemoglobin types except the alkali-resistant hemoglobin F.
The elucidation of the mechanism allowed us to formulate two strategies to extend the interference-free range of this method. The use of a rate-correction method (similar to that used in the creatinine/rate-blanked method) should permit the method to be useful up to a HBOC concentration of 30 g/L depending on the initial alkaline phosphatase activity. However, an even more promising approach may be to use 450 (or 440) nm as the analytical wavelength, which based on our initial studies would appear to be free from interference up to a HBOC concentration of 50 g/L.
In general, it is expected that the majority of blood substitute-containing specimens will have a concentration <50 g/L. In situations such as trauma where many units of HBOC might be transfused, the use of alternative wavelengths in the correction of test interferences is likely to be the preferred approach because it avoids the excessive absorbance values associated with the characteristic wavelengths of hemoglobin. In addition, because it was demonstrated that the magnitude of the interference is a function only of the concentration of blood substitute and is independent of the alkaline phosphatase activity, the proposed corrections should also be valid for increased values of alkaline phosphatase such as observed in the pediatric population.
It is clear from these studies that understanding the underlying mechanism can be an indispensable tool in the development of new or modified methods that have reduced interference from blood substitutes. In the case of the alkaline phosphatase method and probably other methods, mechanisms other than the commonly cited, mere superimposition of a hemoglobin-like absorbance spectrum can be involved. Manufacturers of clinical chemistry analyzers will need to perform complete interference studies involving large amounts of data to document the performance of new or modified methods at all anticipated concentrations of both analyte and blood substitute. It is hoped that the results presented here will serve as a solid starting point for manufacturers in resolving the strong interference seen in the alkaline phosphatase method. The experiments described herein may also serve as a model for the investigation and resolution of blood substitute interferences in other tests.
The results presented herein should also be of great interest to manufacturers interested in improving the hemolysis index for the alkaline phosphatase method. Because the mechanism of interference is identical for both native and polymerized hemoglobin, the correction strategies presented will be applicable to efforts to increase the extent of hemolysis permitted (i.e., hemolysis index) for reporting of results. This would be a small but welcome improvement in the clinical laboratory, regardless of whether blood substitutes are in use.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Nonstandard abbreviations: HBOC, hemoglobin-based oxygen carrier; AMP, 2-amino-2-methyl-1-propanol; and UV, ultraviolet.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. S. Jahr, S. Osgood, S. J. Rothenberg, Q.-L. Li, A. W. Butch, R. Gunther, A. Cheung, and B. Driessen Lactate Measurement Interference by Hemoglobin-Based Oxygen Carriers (Oxyglobin(R), Hemopure(R), and HemolinkTM) Anesth. Analg., February 1, 2005; 100(2): 431 - 436. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and product of alkaline
phosphatase assay (4-nitrophenol; 
).
, 60 U/L.
, 10 g/L HBOC only.
