Clinical Chemistry 44: 2277-2280, 1998;
(Clinical Chemistry. 1998;44:2277-2280.)
© 1998 American Association for Clinical Chemistry, Inc.
|
Enzymes and Protein Markers |
HPLC detection of fetal blood in meconium: improved sensitivity compared with qualitative methods
Dan Chen1,
Timothy R. Wilhite2,
Carl H. Smith1,2,
Morey A. Blinder1,
and Michael Landt1,2,a
Departments of
1
Pathology and
2
Pediatrics, Washington University School of Medicine, St. Louis, MO 63110.
a Address correspondence to this author at: Department of Pediatrics, Washington University School of Medicine, One Children's Place, St. Louis, MO 63110. Fax 314-454-2274; e-mail landt{at}kids.wustl.edu.
 |
Abstract
|
|---|
We describe an HPLC-based method for the detection and quantification of
fetal hemoglobin in stools of newborns. The new procedure is an
alternative to the classic qualitative test for adult hemoglobin in
meconium based on the differential stability of hemoglobin species in
dilute base (Apt test). The HPLC method, based on a commercial device
for hemoglobin characterization (Bio-Rad Variant), readily separates
fetal and adult hemoglobin from non-hemoglobin components of meconium.
To validate the method, blood and meconium were mixed in various
proportions and then prepared for analysis with extraction in saline.
The HPLC method accurately identified hemoglobin species even when the
blood constituted only 5 mL per 100 g of the meconium specimen,
and nearly quantitative recovery of hemoglobin was obtained at a blood
content of 20 mL per 100 g of the meconium. Analysis time was 6.5
min, and preparation of sample was simple. HPLC detection of fetal
blood in stools or other specimens markedly improves
detection/characterization of blood in meconium.
|
Introduction
|
|---|
The Apt test is a qualitative test to identify the source of blood
present in stool of newborns for differential diagnosis of
gastrointestinal bleeding (1). The principle of the test is
based on the different susceptibilities of hemoglobin A (HbA, the major
component of adult hemoglobin) and hemoglobin F (HbF, the major
component of fetal/newborn hemoglobin) to alkaline denaturation. The
latter is more resistant to alkaline denaturation than the former;
therefore, samples containing visually detectable HbF retain a pink
color, whereas samples containing HbA fade to a yellowish color after
alkaline treatment. Because newborn blood contains 65–90% HbF and
adult blood usually usually contains <1%, knowledge of whether HbF is
present differentiates between newborn gastrointestinal bleeding and a
maternal source of the blood (2). In general, grossly bloody
stool specimens of newborns are required for accurate analysis because
the outcome of the test is determined visually; the requirement for
visible blood limits sensitivity, and the subjective nature of
detection introduces operator-dependent variation.
To overcome these problems, we developed an HPLC method to identify the
species of hemoglobin in stool specimens. We utilized the Variant HPLC
system (Bio-Rad Corp.), which separates hemoglobin species by
cation-exchange chromatography and quantifies the presence of each type
of hemoglobin (3). Our results indicate that the HPLC method
has much higher sensitivity and specificity than the Apt test.
|
Materials and Methods
|
|---|
stool specimens
Meconium from newborns (0–48 h of age) was pooled and stored at
-20 °C until the day of the experiment. In the standard method, a
weighed amount of meconium (0.05–0.08 g) was mixed with saline (9 mL
of saline per gram of meconium), and the meconium was thoroughly
suspended by vortex mixing. The specimen was frozen and thawed to lyse
cells. After centrifugation of the mixture at 8800g for 10
min, the supernatant was used for analysis. In validation experiments,
cord or adult blood was added before addition of saline to simulate a
bloody stool. Paired controls for each stool specimen were prepared by
substituting an equal volume of saline solution for meconium. Adult
blood was drawn from healthy adults into EDTA-containing tubes; whole
blood was stored at 4 °C for a maximum of 2 days before use. Cord
blood was obtained as remainders of EDTA-anticoagulated specimens sent
for blood typing to the hospital blood bank. Cord blood samples were
stored at 4 °C for up to 48 h before use. This study was
conducted in accordance with protocols approved by the Human Studies
Committee of Washington University.
apt test
The procedure was performed in the laboratory by two experienced
technologists. A small amount of specimen was spotted onto paper, and
the specimen area on the paper towel was saturated with one drop of 250
mmol/L NaOH. An immediate color change (from pink to brown) indicated
the presence of HbA; persistence of a pink color indicated the presence
of HbF. Adult blood and cord blood served as negative and positive
controls, respectively.
hplc analysis
HPLC analysis was performed with a Bio-Rad Variant analyzer, using
the manufacturer's reagents and following the manufacturer's protocol
(ß-thalassemia short program). The analyzer was programmed with a run
time of 6.5 min, column temperature of 30 ± 4 °C, and
two-buffer gradient elution system using sodium phosphate buffers
(overall flow rate, 2.0 mL/min).
|
Results
|
|---|
hplc detection of hemoglobin in meconium
Because the HPLC method was designed for analysis of red cell
lysates and the application sought was in meconium/stool, it was
necessary to determine whether the method could detect HbA and HbF in
the presence of meconium. Meconium contains considerable pigmented
material, which absorbs at the wavelength used by the Variant (415 nm)
to detect eluting hemoglobin species; when adult blood is present at 50
mL/100 g (500 µL/g) of meconium, >25% of the absorbance at that
wavelength is attributable to soluble meconium pigments. A mixture of
adult blood, fetal blood, and meconium was subjected to analysis after
addition of saline and production of a supernatant by centrifugation.
The hemoglobin fractions in the specimen were well separated by the
HPLC, and non-hemoglobin components that contributed to absorbance
eluted well before the hemoglobin species (Fig. 1
). The chromatogram showed two major peaks with elution times
corresponding to HbF and HbA, identified by retention times retained by
the HPLC software and validated by analysis of adult and newborn blood
specimens.
View larger version (27K):
[in this window]
[in a new window]
|
Figure 1. HPLC separates HbA and HbF in the presence of meconium.
A mixture of equal parts of cord and adult whole blood was mixed with
meconium (20 mL/100 g), and extraction was performed as described in
Materials and Methods. Peaks 1,
2, and 3 represent elution of meconium, HbF, and
HbA, respectively.
|
|
hplc quantification of hemoglobin in meconium specimens
Because meconium might interfere in the measurement of hemoglobin
by HPLC, we investigated the recovery of HbA and HbF in
meconium-containing samples. The recovery of total hemoglobin detected
in meconium samples was compared with that in paired controls where the
meconium was replaced with saline. Identification of HbF required the
presence of only 5 mL of cord blood in 100 g of meconium (50 µL
in 1.0 g). When the ratio of cord blood to meconium was >20
mL/100 g (200 µL/g), recovery was consistently 85% of that in saline
controls (Fig. 2
A). Recovery of adult hemoglobin was similar to that of fetal
hemoglobin (Fig. 2B
).
View larger version (13K):
[in this window]
[in a new window]
|
Figure 2. The recovery of total hemoglobin from meconium specimens
with added cord (A) or adult (B) blood.
Blood was mixed with meconium at different volume:weight ratios,
starting at 5 mL/100 g. In control samples, saline replaced meconium.
The percent recovery is the quantity of hemoglobin detected in the
meconium sample divided by that in the paired control. Each value was
the mean of results from duplicate samples.
|
|
hemoglobin stability in meconium
Adult hemoglobin in meconium (20 mL/100 g), after addition of
saline for the usual specimen processing but before centrifugation, was
stable up to 8.5 h when stored at room temperature and for 24
h at 4 °C (Fig. 3
A). Fetal hemoglobin in meconium was similarly stable (Fig. 3B
).
Addition of EDTA and/or aprotonin did not appear to improve stability
(data not shown).
View larger version (15K):
[in this window]
[in a new window]
|
Figure 3. Stability of hemoglobin in meconium samples.
Cord (A) or adult (B) blood was mixed with
meconium (20 mL/100 g). Hemoglobin was measured at different time
points after storage at room temperature ( ) or 4 °C ( ), and
the total hemoglobin detected was divided by that of paired controls in
which meconium was replaced with saline.
|
|
comparison of hplc and apt methods
Duplicate meconium specimens with added cord blood were analyzed
simultaneously with both methods. The Apt test required a minimum of 40
mL/100 g cord blood (with HbF 80% of the total hemoglobin) for
detection of fetal hemoglobin (Table 1
). Moreover, the reproducibility of the Apt test was very poor.
Even with specimens containing 60–80 mL/100 g cord blood, detection of
a positive visual endpoint was inconsistent. This inconsistency is
likely caused, at least in part, by the difficulty in appreciating the
red color contributed by hemoglobin in the presence of meconium, which
made detection difficult with the visual Apt methodology. Generally,
specimens containing <40 mL/100 g hemoglobin did not have a detectable
red color to base analysis on. In contrast, the HPLC method required
only 5 mL of cord blood per 100 g of meconium to identify the
hemoglobin species present reproducibly (Table 1
), despite reduced
quantitative recovery at this concentration.
precision
Repeated analysis of a specimen with 500 µL of whole blood added
(equal volumes of cord and adult blood) per gram of meconium yielded a
day-to-day imprecision (CV) of 4.3% for the peak area for total
hemoglobin and 3.8% for the peak area of HbF (n = 8).
|
Discussion
|
|---|
Here we describe the development of a new method to detect the
presence of HbF in newborn meconium. Traditionally, the Apt test
(1) has been used to distinguish the source of blood in such
specimens for differential diagnosis of newborn gastrointestinal
bleeding. Identification of HbA as the predominant hemoglobin species
in the specimen indicates that the blood source is of maternal origin,
likely as the result of ingestion of blood in the amnion or at birth. A
finding of maternal origin for the blood can spare the newborn infant
from having to undergo unnecessary procedures such as radiographic
examinations (1). The dependence of the Apt test on
subjective visual judgment for analysis yields low sensitivity and poor
reproducibility, especially when the meconium specimen is not grossly
bloody. Earlier investigations sought to modify the Apt test to improve
the reliability of distinguishing HbA and HbF by measuring HbF-specific
absorbance at 576 nm (4)(5). In one study, HbF
was linearly proportional to absorbance, which allowed for
quantification (4).
The HPLC method we describe here has advantages over both the original
Apt test and the modified procedures based on alkaline denaturation
(1)(4)(5). It is a rapid, automated,
and quantitative method with high sensitivity and specificity. It
eliminates the dependence on subjective visual interpretation of the
Apt test and is capable of detecting fetal hemoglobin in the presence
of meconium at concentrations far below the detection limit of the Apt
test. The Apt test is not widely available, probably because of
infrequent demand combined with the difficulty of maintaining specific
reagents and proficiency for an infrequently used test. Because the
HPLC test utilizes equipment and reagents already present in many
laboratories and used for HbA1c or hemoglobin species analysis,
maintaining the HPLC test will be much more feasible. Improved
feasibility may increase the availability of this test.
Specimens prepared for HPLC analysis are sufficiently stable that
analysis may be delayed for up to 24 h if permitted by the
clinical circumstance. Thus, it is possible to transport specimens to a
reference laboratory or another hospital for analysis when the test is
not available on-site. The procedure is sufficiently rapid (preparation
and analysis <1 h) to allow emergency analysis in critically ill
infants. Finally, the HPLC method eliminates the need for a caustic
reagent (sodium hydroxide).
The utility of hemoglobin analysis in meconium and stool may be greatly
expanded by the increased sensitivity of HPLC, which will allow serial
measurement to follow resolution of gastrointestinal bleeding with
therapy. Moreover, it may be useful in other clinical applications to
distinguish maternal and newborn blood, such as in cordocentesis
sampling or in body fluids such as amniotic fluid or vomitus
(6)(7)(8). Because the method provides quantitative assessment
of the fraction of each hemoglobin species present, it is possible to
determine relative contributions of maternal and fetal blood in
situations of mixed blood sources; this determination is aided by the
ease of quantifying HbF percentages in the blood of the mother and the
newborn.
|
Acknowledgments
|
|---|
We commend Michael J. Lukoszyk and Eva E. Reyes for excellent
technical assistance and Barbara Hartman for assistance in the
preparation of the manuscript.
|
References
|
|---|
-
Apt L, Downey WS, Jr. "Melena" neonatorum: the swallowed blood syndrome. J Pediatr 1955;47:6-12.
[ISI][Medline]
[Order article via Infotrieve]
-
Kim HC, Schwartz E. Fetal hemoglobin. Beutler E Lichtman MA Collen BS Kipps TJ eds. Williams hematology 5th ed. 1995:L42-L43 McGraw-Hill New York. .
-
Papadea C, Cate JC, IV. Identification and quantitation of hemoglobins A, F, S and C by automated chromatography. Clin Chem 1996;42:57-63.
[Abstract/Free Full Text]
-
Liu N, Wu AHB, Wong SS. Improved quantitative Apt test for detecting fetal hemoglobin in bloody stools of newborns. Clin Chem 1993;39:2326-2329.
[Abstract]
-
Crook M. Hemoglobin in stools from neonates: measurement by a modified Apt test. Med Lab Science 1991;48:346-347.
-
McRury JM, Barry RC. A modified Apt test: a new look at an old test. Pediatr Emerg Care 1994;10:189-191.
[ISI][Medline]
[Order article via Infotrieve]
-
Ogur G, Gul D, Ozen S, Imirzalioglu N, Cankus G, Tunca Y, et al. Application of the Apt test in prenatal diagnosis to evaluate the fetal origin of blood obtained by cordocentesis: results of 30 pregnancies. Prenat Diagn 1997;17:879-882.
[ISI][Medline]
[Order article via Infotrieve]
-
Mori M, Holbrook RH, Jr. Application of the hemoglobin alkaline denaturation test to determine the fetal origin of blood: applicability to funipuncture. Obstet Gynecol 1993;81:793-796.
[ISI][Medline]
[Order article via Infotrieve]
Top
Abstract
Introduction
