USA HCG REFERENCE
SERVICE
HYPERGLYCOSYLATED HCG
Click on link below to show the most
recent publications-
NEW REVIEW
Cole LA, Hyperglycosylated hCG, Placenta, 2007; in press.
Cole
LA, Dai D, Leslie KK, Butler SA, Kohorn EI.
Gestational trophoblastic diseases: 1.
Pathophysiology of hyperglycosylated hCG-regulated neoplasia.
Gynecologic Oncology, in press, 2006
Cole LA, Khanlian SA, Riley JM,
Butler SA. Hyperglycosylated hCG (hCG-H) in Gestational Implantation,
and in Choriocarcinoma and Testicular Germ Cell Malignancy Tumorigenesis,
J Reprod Med, in press 2006
Laurence
A. Cole, Sarah A. Khanlian, Jaime M. Riley
and Stephen A. Butler
Hyperglycosylated
hCG Review
Cole LA and Khanlian SA
USA hCG Reference Service
Hyperglycosylated human chorionic gonadotropin, acronym hCG-H,
has been shown to be a completely independent molecule to regular hCG,
produced by separate cells and having a totally different function
(1-4). It is for this reason that we call it a biologically independent
molecule to regular hCG. Its established functions are, however,
autocrine rather than endocrine, more like those of a cytokines (1-8).
Hyperglycosylated hCG has double size O-linked sugar structures and
larger N-linked structures than hCG, changing its biological function
(5-8). Here we review hCG-H biochemistry and clinical chemistry. We
review research on the following topics: a. Structure of hCG-H; b.
Source and Biological Function of hCG-H; c. Measurement of hCG-H; d.
hCG-H in Pregnancy Detection; e. hCG-H Measurement
in Detection of Pregnancy Outcome; f. hCG-H in Prediction of
Gestation Down Syndrome; g. hCG-H in Management of Gestational
Trophoblastic Diseases.
a. Structure of hCG-H
hCG is a heterogeneous molecule
produced by trophoblastic cells in pregnancy. Peptide variants are
detectable in serum and urine samples during pregnancy and gestational
trophoblastic diseases (9). Oligosaccharide variants originate from
differences in the availability of sugars, variations in cellular
metabolism, and the differential expression of different
glycosyltransferases, the sugar-adding enzymes in cells (5, 8, 10). It
has long been recognized that the hCG molecule produced in
choriocarcinoma or cancer of the trophoblastic cell is a larger molecule
than that produced through most of normal pregnancy [11-13]. In 1987-8,
a significant and consistent difference was demonstrated between the 4
O-linked oligosaccharides on hCG from choriocarcinoma patients and those
found on hCG from normal pregnancy (6, 7). This was later confirmed in a
study with a larger number of pregnancy and choriocarcinoma cases in
1997 (5).
While the principal O-linked
oligosaccharide structure in all normal pregnancies is a trisaccharide
with the structure NeuAc2-3Galß1-3GalNAc-O-Serine (where NeuAc is sialic
acid, Gal galactose and GalNAc is N-acetylgalactosamine), the principal
structure in all choriocarcinoma cases is a double size hexasaccharide
with the structure
NeuAc2-3Galß1-3(NeuAc2-3Galß1-4GlcNAcß1-6)GalNAc-O-Serine (where GlcNAc
is N-acetylglucosamine) (5-7). In 1997 (5) we demonstrated that the
difference in the 4 O-linked oligosaccharides is the principal
difference between choriocarcinoma and pregnancy hCG. While first
trimester normal pregnancy urine hCG contained 12.3 to 19% of the larger
hexasaccharide side chains (n=6 individuals, mean = 15.6%),
choriocarcinoma urine hCG contained 48 to 100% hexasaccharide side
chains (n=6 individuals, mean = 74.2%).
Lesser differences have also been shown
in the 4 N-linked oligosaccharide structures on pregnancy and
choriocarcinoma molecules (5, 8, 10). While in normal pregnancy,
sialyl-N-acetylactosamine biantennary N-linked oligosaccharides and
fucosylated N-linked oligosaccharides predominate at the two sites on
the ß-subunit of hCG in normal pregnancy, significantly larger
fucosylated sialyl-N-acetylactosamine triantennary oligosaccharides
predominate in choriocarcinoma cases (5,8). While in normal pregnancy
sialyl-N-acetylactosamine biantennary N-linked oligosaccharides and
biantennary oligosaccharides with the
a1,3
antenna terminating in Mannose predominate at the two sites on the
a-subunit
of hCG in normal pregnancy, significantly larger fucosylated biantennary
oligosaccharides predominate in choriocarcinoma cases together with
abnormal triantennary oligosaccharides with the
a1,3
antenna terminating in Mannose (5,6).
The oligosaccharides on hCG in normal
pregnancy contribute to the molecular weight of 36,700 daltons.
Choriocarcinoma hCG, in contrast, with its large sugar side chains has a
molecular weight of approximately 41,000 daltons (5). The name
hyperglycosylated hCG (hCG-H) was formulated for this molecule with
unduly large sugar side chains.
b. Source and
Biological Function of hCG-H
hCG with larger O-linked
oligosaccharides (hCG-H) characterizes the molecules produced in
choriocarcinoma. A specific monoclonal antibody to hCG-H (antibody B152)
was generated against the hCG with 100% hexasaccharide O-linked
structures (100% hCG-H) produced by a single patient with
choriocarcinoma (14). Using this monoclonal antibody specific sandwich
assays have been established for detecting
intact hCG-H (4, 15).
Multiple publications
examining hCG carbohydrate structure shows that patients with
choriocarcinoma and choriocarcinoma cell line produce primarily a
molecule with primarily hexasaccharide O-linked oligosaccharide and an
abundance of triantennary N-linked oligosaccharides, or
hyperglycosylated hCG (5-8). This has been demonstrated now in large
numbers of choriocarcinoma or gestational trophoblastic neoplasm (GTN)
cases using the microtiter plate and commercial B152-based hCG-H assays
(16, 17).
In 1998 O’Connor et
al. used a B152-based microtiter plate assay to show that hCG-H is
not only the principal form of hCG made in choriocarcinoma, but also the
principal form of hCG made during the initiation of pregnancy, at the
time of and in the 2 weeks following implantation of pregnancy (18).
This finding has now been confirmed by multiple investigators (3, 4,
14-20). Recent data on the proportion of total hCG accounted for by
hCG-H in urine samples is presented in Table 1 (16-20). A similarity
has previously been suggested in the invasion processes of implantation
and choriocarcinoma (21, 22)
Cytotrophoblasts are phenotypically invasive cells. These are the
principal cells in choriocarcinoma and GTN, the active cells in
choriocarcinoma tumors, and the active invasive cells of blastocysts at
the time of implantation (21-23). As published in multiple articles
between 1999 and 2005, the cytotrophoblast cells of pregnancy
trophoblast and of choriocarcinoma cell lines produce hCG-H, and the
differentiated syncytiotrophoblast cells of pregnancy produce regular
hCG (1, 3, 4). As such the high proportion of hyperglycosylated hCG of
implantation and of choriocarcinoma is produced by cytotrophoblast
cells. The hCG produced in pregnancy is made in syncytiotrophoblast or
different cells.
The primary
function of hCG in pregnancy is to maintain progesterone production by
corpus luteal cells (24). In contrast, hCG-H is only associated with
invasive events, pregnancy implantation, and GTN or choriocarcinoma (5).
The choriocarcinoma hCG immunoreactivity in the conditioned medium of
JAR choriocarcinoma cells (invasive cytotrophoblast cells) is hCG-H
(16). Studies by Lei et al. (2) in 1999, show that JAR cells are
invasive in Matrigel basement membrane inserts (tumor invasion models)
in vitro, and are rapidly tumorigenic when transplanted into
athymic nude mice in vivo. Lei et al. (2) treated JAR
cells with hCG a-subunit
antisense cDNA. This blocked production of
a-subunit
and thus of hCG dimer in the choriocarcinoma cells (cells primarily
produce hCG-H (1, 16), not specified by Lei et al.) and also
blocked Matrigel membrane insert invasion in vitro and
tumorigenesis in athymic nude mice in vivo. This data indicated
that hCG-H has a separate role to hCG, a critical function in
trophoblast cell invasion. Recently, we demonstrated that hCG-H has
minimal activity at the corpus luteal hCG receptor (1). When isolated
normal pregnancy cytotrophoblast cells and JEG-3 choriocarcinoma cells
(100% cytotrophoblast) were cultured on Matrigel basement membrane
inserts invasion occurred. In both cell sources, invasion was
significantly promoted by hCG-H and not by hCG. JEG-3 choriocarcinoma
cells are rapidly tumorigenic when transplanted into athymic nude mice
in vivo (1). Cole et al. (1) treated transplanted mice
with B152 antibody against hCG-H. Antibody therapy blocked tumor
formation and tumor growth in vivo (1). These data suggest that
hCG-H has a completely separate role to hCG. It has a critical function
in cytotrophoblast invasion or implantation in pregnancy, and in tumor
invasion or tumorigenesis in choriocarcinoma cases. That cytotrophoblast
cells produce hCG-H, and that it acts directly on cytotrophoblast cells
promoting invasion suggesting an autocrine function. That circulating
antibody against hCG-H inhibits the invasive action indicates hCG-H has
an autocrine action.
In summary, hCG-H has a
different structure to hCG, it is produced by separate cells to hCG, and
acts on separate cells to hCG, with a very different tumorigenic
function. As such, it needs to be considered as a completely independent
molecule. It can be considered as a biologically independent molecule to
regular hCG, yet it is not an endocrine or hormone per se. It also is
ineffective in promoting progesterone production, so is not a
gonadotropin like hCG.
c. Measurement of hCG-H
A specific monoclonal antibody to hCG-H
(antibody B152) was generated against the hCG with 100% hexasaccharide
O-linked structures (100% hCG-H) produced by a single patient with
choriocarcinoma (14). Using this monoclonal antibody a microtiter plate
two antibody (B152 plus anti-ß tracer) assay was established for
detecting intact hCG-H (4). Nichols Institute
Diagnostics (San Clemente CA) now markets an automated chemiluminescence
test using antibody B152 and an anti-ß tracer to detect hCG-H, for the
Nichols Advantage immunoassay platform (15). This assay has been FDA
approved for pregnancy-related applications. As published, the
microtiter plate assay with overnight or 4 hours incubation with
antibody B152, and the automate test with 20 minutes incubation have
different specificity (4, 14, 16). The microtiter plate assay recognizes
hCG-H (100% immunoreactivity), sialic acid deficient hCG-H (100%), hCG-H
free ß-subunit (60%), pure CHO-cell recombinant hCG (<1%), and pure CHO
(Chinese Hamster ovary) -cell recombinant hCG free ß-subunit (<1%). The
automated Nichols Advantage assay has similar specificity but
considerable lower affinity for hCG-H sialic acid deficient (~30%) and
free ß-subunit (~20%) variants (4, 14, 16, 25). The difference may be
due to the 20 minute incubation time in the automated assay, compared
with 4 hrs in the microtiter plate test. Equilibrium has been
demonstrated in the later test, but likely does not occur in the
automated test.
As published (3, 4,
14-20), the majority of total hCG immunoreactivity in serum and urine
samples in early pregnancy samples is due to hCG-H. As shown in Table 1,
hCG-H accounts for a 92% (median) of total hCG in the 3rd
complete week of gestation or the week following implantation, 73% in
the 4th week or the week following missing menses when most
pregnancy testing is done, and then declines rapidly thereafter. It
accounts for 14% in the 7th week of gestation and <2% in the
second and third trimesters of pregnancy. Clearly a pregnancy test needs
to appropriately detect hCG-H. As shown in Table 2, only 3 of 14
professional laboratory serum pregnancy tests equally or near-equally
detect hCG and hCG-H (16, 26). The majority (11 of 14 tests), under
detect or very poorly detect (9 of 14 tests) hCG-H (16). Similarly, only
4 of 15 home pregnancy tests or over the counter tests equally detect
hCG and hCG-H (16). None of the tested point of care or physician’s
office tests equally detect hCG and hCG-H (16). Again, the vast majority
of test (11 of 15 over the counter and 2 of 2 point of care) very poorly
detect hCG-H (16, 20). Over the counter or point of care devices make
claims like “99% accurate” or “100% accurate,” and “Use as Early as the
First Day of Missed Period” or “Use Test Before You Expect Your Period.”
Claims like these are concerning, especially when the device does not
appropriately detect the principal hCG analyte in early pregnancy
samples. One might ask with all these professional laboratory and front
line pregnancy testing point of care and over the counter devices, are
they really pregnancy tests if they poorly detect the principal analyte
that marks early pregnancy?
Today, 2005, 18
years since the literature first described hCG-H, and 7 years since
publications started to show that this is the principal hCG
immunoreactivity produced in early pregnancy, most manufacturers
continue to ignore hCG-H and don’t make assays optimized for detecting
hCG-H. This is primarily due to the non-availability of a formal
standard or W.H.O standard. In recent years, several manufacturers have
produced their own standard to optimize their assays. These standards
have been prepared by partially purifying the hCG-H in JEG-3
choriocarcinoma cell line culture fluid (16). The hCG-H produced by JaR
and JEG-3 (clones) culture fluids has been characterized and shown to be
similar to choriocarcinoma patient hCG-H (16, 27). This culture fluid
hCG-H has been calibrated using the Nichols Advantage hCG-H test
(calibrated against a JEG-3 cell line standard). It appears that most
manufacturer’s will follow FDA guideline of calibrating tests against
hCG alone until either guidelines are changed or a WHO hCG-H standard is
produced. Until that time mostly sub-optimal pregnancy tests with poor
recognition of hCG-H may be continue to be produced.
d. hCG-H in Pregnancy
Detection
As indicated in Table 1, hCG-H is a
marker of early pregnancy. Measurement of both hCG and hCG-H equally is
optimal for hCG pregnancy testing. Between one third and one quarter of
pregnancies may fail hours or days after implantation (18, 28). There
are considered biochemical pregnancies or early pregnancy losses. When
monitoring natural and in-vitro fertilized it is important to try
and distinguish “clinical” pregnancies which may make it through to term
from “biochemical” pregnancies which fail to thrive. As published (18,
28), all biochemical pregnancies fail to produce hCG-H. The lack of
hCG-H or the invasive stimulator could actually be the cause of failure.
Ongoing studies in our laboratory, monitoring women eager to achieve
pregnancy show that approximately 1 in 4 “biochemical” pregnancies
produce no detectable hCG-H and 3 of 4 produce borderline detection
amounts of hCG-H compared to “clinical” pregnancies. hCG-H may
potentially be a better pregnancy test than hCG because it does not or
poorly detects “biochemical” pregnancies (18, 28, ongoing studies in
Cole laboratory).
Women are administered hCG to promote
maturation of multiple follicles for collection for in- vitro
fertilization. As such, the slow disappearance of exogenous hCG has to
be demonstrated, over 2 weeks, before endogenous production from a
pregnancy can be demonstrated. Women who pay for in-vitro
fertilization, and go through the operative procedures are usually eager
to learn about their success or failure in achieving pregnancy. Waiting
for exogenous hCG to clear requires a two week delay. Detection of hCG-H
instead of hCG can avoid this delay. A pregnancy test measuring hCG-H
may have multiple advantages both in the earlier detection of
pregnancies in assisted reproductive technologies, and in the avoidance
of detecting “biochemical” pregnancies.
e. hCG-H Measurement in
Prediction of Pregnancy Outcome
In 2002 Kovalevskaya et al.,
discovered unusually low proportions of hCG-H in pregnancies that
spontaneously abort in the first trimester of pregnancy (28). The low
proportions started when hCG was first detected and continue until the
time of spontaneous abortion. While this was an important discovery no
clinical application was indicated. Two years later these findings were
confirmed by Sutton-Riley et al. (29) and Byrn et al. (30).
Sutton-Riley et al. (29) showed that a
simple single point cut-off of 13 ng/ml hCG-H in serum could be used to
differentiate a failure outcome (<13 ng/ml) from term outcome (>13
ng/ml) pregnancy between 4 and 7 weeks of pregnancy (29). As shown in
Table 3, this 13ng/ml cut-off was pivotal in detecting 73% of failures
(spontaneous abortion and ectopic pregnancy) at 5% error rate, compared
to 42% detection at this same error rate for a comparable cut-off for
hCG. A highly significant difference was observed between the area under
the ROC curve results, or test accuracies using hCG and hCG-H
(P<0.00005). The use of this 13 ng/ml cut-off for detecting failures in
urine was also demonstrated, with the feasibility of a 13 ng/ml marked
point-of-care test for physician’s office applications. Byrn et al.
(30) showed that a hCG-H single test or multi-point doubling test
offered great improvement over an hCG single point or doubling test in
detecting failing pregnancies. They showed that the hCG-H test offered
improvement in both the detection of spontaneous abortion and ectopic
pregnancies.
All 3 independent studies, with
different populations (28-30) supported each other in showing that
unduly low hCG-H accompanies spontaneously aborting pregnancies, from
time of first hCG detection onwards. As discussed earlier (1-2), hCG-H
has a clear role in promoting the invasive process in implantation of
pregnancies. It is speculated that low hCG-H production, or trophoblast
differentiation of the hatched blastocyst at the time of implantation,
may lead to inappropriate implantation, and may be a major cause of
spontaneous abortion of pregnancy.
f. hCG-H in Prediction
of Gestation Down Syndrome
The first application demonstrated for
hCG-H was in screening for gestational Down syndrome.
In 1997 a preponderance of hCG-H was
demonstrated in structure studies in mid-trimester Down syndrome
pregnancies, and then later confirmed by using the B152-based hCG-H
immunoassay (31, 32). This was a surprising finding. Why Down syndrome
pregnancies? This enigma has been explained in recent years by the
finding of a defect in cytotrophoblast differentiation into
syncytiotrophoblast cells in Down syndrome cases (33 -35), or a
resultant accumulation of cytotrophoblasts, the hCG-H producing cells,
in Down syndrome pregnancies.
Clinical trials were initiated in 1997 at Yale University (New
Haven, CT). Serum and urine samples were sought over a period of 2 years
from all women waiting to have amniocentesis because of maternal age
concerns (4). None had been pre-selected with biochemical methods (i.e.
triple or quadruple screen). While all consenting women volunteered a
urine sample (1526 total), only 34% of this group also agreed to
phlebotomy or to providing a serum sample (512 total). Samples were
collected between 14 and 22 weeks of gestation and tested blindly
(before karyotype known). All urine samples were collected from the
amniocentesis center daily stored in a refrigerator (no freezing) and
tested once each week for hCG-H and creatinine. hCG-H was measured using
the B152-based microtiter plates test for hCG-H (4, 36). Tests on spot
urine samples were normalized to creatinine concentration.
MoM (multiples of median) statistics
were used to analyze data. The logarithmic regression equation relating
the hCG-H results for 1487 normal karyotype patients with gestational
age (ga) was median = 6050 x (0.71ga).
The median MoM for normal karyotype patients was 1.0 (standard deviation
0.47), and the median MoM for the 39 later proven Down syndrome cases
was 9.5 (4). These data indicated that Down syndrome samples had
9.5-fold higher hCG-H values than those with normal karyotype. In
addition, it was observed that 31 of 39 Down syndrome cases had hCG-H
results that exceeded the 95th centile of normal karyotype
cases. ROC analysis indicated that this single test would achieve 80%
detection at a 5% false positive rate. A triple test (hCG +
unconjugated estriol + a-fetoprotein)
was later performed on the 512 serum samples provided, 64% detection of
Down syndrome cases was observed at 5% false positive rate. Multivariate
analysis combining maternal age, urine hCG-H results, and serum triple
test values yielded 96% detection with a 5% false positive rate and 94%
detection at a 3% false positive rate.
In an initial serum study, hCG-H was measured in serum samples
from 10 gestational Down syndrome cases and 66 normal karyotype
pregnancies at 15-22 weeks of gestation. A lesser elevation was noted
with these serum samples. Down syndrome samples had a median of 3.9 MoM.
Using serum hCG-H concentration alone, 60% detection was indicated at a
5% false positive rate (37). In an independent serum study evaluating 19
gestational Down syndrome cases and 82 controls, collected in multiple
centers in Connecticut and blindly tested at Quest Diagnostics, 81%
detection was observed with the hCG-H test at a 5% false positive rate.
In a multivariate analysis, combining serum hCG-H with a quadruple test
(hCG + unconjugated estriol + a-fetoprotein
+ inhibin), 95% detection was achieved at a 5% false positive rate or
92% detection at a 3% false positive rate (38). In a further study,
looking at 49 Down syndrome cases shipped at ambient temperature and
stored at -20°C for 8 years, only 55% detection at 5% false positive was
indicated (39).
The high sensitivity of 80% and high 9.5 median MoM was
reported with the multititer plate assay with long incubation with B152
(4). In similar size studies with serum and urine samples using the
Nichols Advantage hCG-H tests yielded much poorer detection rates (50%
detection at 5% false positive rate) and lower elevations in Down
syndrome cases (3.5 MoM) (52-55). This is seeming due to the difference
in specificity of the microtiter plate assay and the automated test with
short incubation. The poor detection of the free ß-subunit of hCG-H
continuously released from dissociating hCG-H, and the poor detection of
the sialic deficient isoforms like explains this difference. This is
probably particularly pertinent when testing frozen libraries of serum
samples, and serum samples shipped by mail as investigated by Pandian
et al (53-55) and by others (40, 42). Recent studies by Sutton et
al (25) clearly demonstrate that the principal isoform of hCG-H
elevated in Down syndrome cases, whether first trimester or second
trimester is hCG-H deficient in sialic acid. A major isoform of hCG-H
was poorly detected by the Nichols Advantage hCG-H test. This explains
the poorer utility of the Nichols Advantage hCG-H test in Down syndrome
screening. We eagerly await a new, fixed or corrected hCG-H test for
this application.
As indicated in Table 4 (4, 16, 20), individual hCG and hCG-H
results vary widely throughout pregnancy. The biggest variation occurs
at the end of the first trimester, when hCG and hCG-H concentrations
rapidly decline following their peak at 8-12 weeks. This wide variation
may be due to inaccuracies in determination of gestational age during a
time when concentrations are plunging every day. More consistent results
are observed in the second trimester (15-22 weeks) when concentrations
are leveling. For this reason, neither hCG nor hCG-H can be expected to
be as good of a screening test at 9-13 weeks as they are in the second
trimester of pregnancy.
In 2000 Weinans and colleagues (40)
blindly tested 136 urine samples from the 10th and 11th
weeks of gestation, including 11 pregnancies shown by chorionic villous
sampling to have Down syndrome fetuses. In that study with the hCG-H
test, 37% detection was observed at 5% false positive rate. This is
similar to the detection rates when using either Pregnancy-Associated
Plasma Protein-A (PAPP-A) or hCG free ß-subunit at the same time of
gestation (41). In 2001, Strom and colleagues (42) used a case control
set of samples involving 4412 women undergoing chorionic villous
sampling or early amniocentesis at multiple-centers. Parallel urine and
serum samples were available from 17 Down syndrome cases from the first
trimester of pregnancy, and from 5 normal karyotype controls for each
Down syndrome case. While the combination of serum PAPP-A and hCG free
ß-subunit measurements detected 60% of Down syndrome cases at a 5% false
positive rate, the combination of urine hCG-H plus serum PAPP-A and hCG
free ß-subunit detected 81% of Down syndrome cases at a 5% false
positive rate, or 74% at a 3% false positive rate (32). A similar but
much larger study was completed recently in Europe with near identical
results (43). Clearly, there is also a role for hCG-H measurement in
first trimester Down syndrome screening. Based upon the findings of
Sutton et al (25), vastly superior results may be expected for
hCG-H in Down syndrome screening one the specificity limitations of the
assay are fixed, or with the introduction of a test specific for the
sialic acid deficient isoform that marks Down syndrome.
g. hCG-H in the
Management of Gestational Trophoblastic Diseases.
Quiescent gestational trophoblastic disease is a benign or
inactive form of GTN
or
choriocarcinoma, marked by persistent low hCG results, pertsisting for
periods ranging from 3 months to 16 years (45-51). It may also be
considered as a pre-malignant state in that approximately 1 in 5 cases
transform into GTN or choriocarcinoma (44-51). We have reviewed the
histology slides from two cases undergoing surgery for this condition.
In both cases, intermediate and highly differentiated
syncytiotrophoblast cells were observed, with a clear absence of
cytotrophoblast cells that mark most cases of choriocarcinoma (45, 46).
Differentiated syncytiotrophoblast cells can be slow growing cells, and
not invasive cells. It is inferred that the presence of
syncytiotrophoblast cells with the absence of invasive cytotrophoblast
cells is the nature of quiescent gestational trophoblastic disease. In
the absence of cytotrophoblast cells, the hCG-H producing cells, no
hCG-H can be detected. That is, syncytiotrophoblast cells remaining
after treatment of an ectopic pregnancy, or parturition or abortion, or
following evacuation of a hydatidiform mole. Alternatively, slow
growing syncytiotrophoblast cells can remain after chemotherapeutic
resolution of cytotrophoblasts or following treatment of GTN or
choriocarcinoma, where these cells may persist.
The USA hCG Reference Service has consulted on 69 cases with
persistent low “real” (not false positive) hCG results (44). In all
cases the persistent low levels persisted for 3 month or greater with
minimal fluctuation and no clear upward or downward hCG trend. hCG-H was
measured using the Nichols Advantage test as a proportion of total hCG
in the DPC Immulite test (% hCG-H) as a marker for invasive
cytotrophoblast cells. In 67 of the remaining 69 cases, which we class
as quiescent gestation trophoblastic disease (44-51), no hCG-H was
detected whatsoever. In the two remaining cases, 10% and 12% hCG-H was
detected. In both, detectable hCG persisted for greater that 6 months
with minimal fluctuation. Considering these histories, and that in these
exceptional cases the proportion of hCG-H was below those of all 79 GTM
or choriocarcinoma cases referred to the Service, the diagnosis of
quiescent gestational trophoblastic disease was conferred. This data
shows the utility of hCG-H as a marker for differentiating quiescent and
invasive disease. Considering that hCG-H is only produced by
cytotrophoblast cells (3-4), the absence of hCG-H in quiescent
gestational trophoblast disease confirms its origin from
syncytiotrophoblast in the absence of cytotrophoblast cells.
Of the 69 identified cases, 33 followed evacuation of a
hydatidiform mole, 15 followed successful treatment of GTM or
choriocarcinoma, and 21 following a spontaneously aborted, ectopic or
term pregnancy. We examined the medical records of these 69 identified
quiescent gestational trophoblastic disease cases. While 42 had received
chemotherapy, combination chemotherapy, hysterectomy or other surgery
for assumed active disease, none fully responded to the therapy,
including chemotherapy. We have interpreted this to mean that theses
cases have slowly growing dispersed syncytiotrophoblast cells, that are
too slow growing to respond to chemotherapy. In all treated cases,
hysterectomy partially but never completely suppressed the hCG result.
We have interpreted this to mean that syncytiotrophoblast cells commonly
remain outside of the uterus after a gestational event or,
alternatively, are transposed through the fallopian isthmus by an
endometriosis-like mechanism. With a 100% summary of records indicating
that therapy does not work, and similar findings by others (50,51), it
is inferred that treating physicians should refrain from using
chemotherapy or surgery in these cases.
Our history with quiescent gestational
trophoblastic disease cases is like that with false positive hCG cases-
too many people receiving needless therapy for a poorly understood and
only recently fully recognized condition that in fact, requires no
therapy. In one case, a patient was shown by the Service on two separate
occasions during the course of one year to have quiescent disease.
Despite our recommendations, additional combinations of chemotherapeutic
agents were given. This patient tragically died from complications of
pulmonary fibrosis following bleomycin chemotherapy. Nevertheless, this
case is important because it emphasizes the importance of making an
accurate diagnosis of GTN/choriocarcinoma versus quiescent disease
before initiating cytotoxic chemotherapy. This requires all medical and
gynecologic oncologists, and to some extent, whomever manages patients
who undergo hCG-H determinations, to be able to recognize and diagnose
the phenomenon of quiescent gestational trophoblastic disease. If the
diagnosis is in doubt or suspect, these data indicate the importance of
referring such patients to centers with experience with this condition
before therapy is initiated.
Published data show the value of hCG-H not just in the
detection of quiescent disease, but in the earlier detection and
differention of recurrent GTN/choriocarcinoma and in the staging of
GTN/choriocarcinoma (44).
Summary
hCG-H is
biologically independent molecule to regular hCG, produced independently
by separate cells and with a completely independent function. hCG-H
offers clinical chemistry a new and extended scope beyond hCG in
management of pregnancy, predicting pregnancy outcome, management of
gestational trophoblastic diseases, and in screening for Down syndrome
pregnancies.
Table 1. The
proportion of total hCG immunoreactivity in urine samples due to hCG-H
in early pregnancy (16, 20). Total hCG was measured in the DPC Immulite
total hCG test, which detects hCG and hCG-H equally (26). hCG-H was
measured in the Nichols Advantage hCG-H test, and hCG-H ng/ml was
expressed as a percentage of total hCG. Gestational age is presented as
complete weeks, where the 3rd complete weeks is 3 weeks 0
days to 3 weeks 6 days since last menstrual period, or in most cases,
the week following implantation.
|
Gestational age |
n |
Median
percentage hCG-H |
Range of hCG-H |
|
|
|
|
|
|
3rd
complete week |
75 |
92% |
55 - 100% |
|
4th
complete week |
63 |
73% |
15 - 100% |
|
5th
complete week |
45 |
50% |
10 - 100% |
|
6th
complete week |
23 |
26% |
7.9 - 53% |
|
7th
complete week |
22 |
14% |
2.3 - 21% |
Table 2. Sensitivity of commercial hCG
tests. Pure intact hCG (CHO cell recombinant hCG, 0% hCG-H) and pure
hCG-H (Preparation C5, 0% hCG, (5)) were calibrated by absorbance at
A278 (26). The concentrations were then assessed blindly in 14
Professional Laboratory Tests (16, 26). Percentages are the assay value
divided by the absolute concentration. hCG-H sensitivity was the
relative sensitivity for detecting hCG-H vs. hCG. Point of care and
over the counter pregnancy test devices were tested for sensitivity by
evaluating using multiple concentrations of absolute calibrated hCG and
hCG-H in at least sextuplicate (16, 20). Sensitivity results are
presented. These are the lowest concentration at which all devices were
positive. hCG-H sensitivity was the relative sensitivity for detecting
hCG-H vs. hCG.
|
Test or Device |
hCG only |
hCG-H only |
hCG-H sensitivity |
|
|
|
|
|
|
|
|
|
|
|
A. PROFESSIONAL LABORATORY TESTS |
|
|
|
|
|
|
|
|
|
Roche Elecsys |
76% |
167% |
220% |
|
Tosoh A1A600 |
95% |
97% |
102% |
|
DPC Immulite |
102% |
99% |
97% |
|
Dade Stratus |
89% |
85% |
96% |
|
Ortho Vitros
ECi |
89% |
80% |
91% |
|
Bayer Centaur |
121% |
107% |
89% |
|
Wako hCG-CTP |
86% |
75% |
87% |
|
Bayer ACS180 |
119% |
102% |
86% |
|
London
radioimmunoassay |
11`2% |
95% |
85% |
|
New Haven
radioimmunoassay |
107% |
91% |
85% |
|
Abbott AxSym |
110% |
86% |
79% |
|
Dade Opus |
90% |
69% |
68% |
|
Dade Dimension |
90% |
50% |
54% |
|
Serono
MAIAclone |
94% |
4% |
4% |
|
|
|
|
|
|
B. POINT OF
CARE TESTS |
|
|
|
|
Beckman Icon
25 |
25 IU/L |
50 IU/L |
50% |
|
Quidel Quick Vue hCG Combo |
25 IU/L |
50 IU/L |
50% |
|
|
|
|
|
|
C. OVER THE
COUNTER TESTS |
|
|
|
|
Answer, |
12.5 IU/L |
12.6 IU/L |
100% |
|
Equate |
12.5 IU/L |
12.5 IU/L |
100% |
|
First
response, Early Result |
9.4 IU/L |
9.4 IU/L |
100% |
|
K-Mart,
American Fare, Easy to Read |
25 IU/L |
25 IU/L |
100% |
|
Clear Blue
Easy, One Minute |
12.5 IU/L |
25 IU/L |
50% |
|
Clear Plan
Easy |
25 IU/L |
50 IU/L |
50% |
|
E.P.T. |
12.5 IU/L |
25 IU/L |
50% |
|
Eckerd,
One-Step |
50 IU/L |
100 IU/L |
50% |
|
Inverness
Medical, Early |
25 IU/L |
50 IU/L |
50% |
|
Inverness
Medical, Early, Cassette |
12.5 IU/L |
50 IU/L |
50% |
|
Save-on, Osco,
One-Step |
25 IU/L |
50 IU/L |
50% |
|
Confirm |
6.3 IU/L |
25 IU/L |
25% |
|
Rite Aid,
One-Step |
25 IU/L |
100 IU/L |
25% |
|
Target Brand,
One-Step |
25 IU/L |
100 IU/L |
25% |
|
Walgreens, One
Step |
12.5IU/L |
100 IU/L |
13% |
|
|
|
Serum hCG-H |
Serum hCG |
|
|
|
|
|
|
Term outcome pregnancies |
n
= 87 |
n
= 87 |
|
Cut-off concentration |
13
µg/L |
125 IU/L |
|
Corresponding false positive rate |
5% |
5% |
|
|
|
|
|
Detection Rates for Failures |
n
= 33 |
n
= 33 |
|
a.
All failures |
24
of 33 (73%) |
14
of 33 (42%) |
|
b.
Spontaneous abortions only |
20
of 29 (71%) |
12
of 29 (41%) |
|
c.
Ectopic pregnancy only |
4
of 4 (100%) |
2
of 4 (50%) |
|
|
|
|
|
|
Predictive value positive |
85% |
76% |
|
|
|
|
|
Area under ROC curve ± SE |
0.88 ± 0.003 1 |
0.71 ± 0.006 1 |
|
ROC 95% confidence interval |
0.83 - 0.99 |
0.78 - 0.97 |
Table 4. Median concentration and
ranges of total hCG concentrations (all forms of hCG including hCG-H) as
measured in the DPC Immulite test and hCG-H concentrations measured in
the Nichols Diagnostics hCG-H assay, in 512 urine samples during the
course of gestation (4, 16, 20). The proportion of ITA was determined as
the ITA result divided by the total hCG result.
|
Gestational age |
n |
total hCG |
hCG-H |
Proportion hCG-H |
|
(weeks) |
|
ng/ml |
Range |
ng/ml |
Range |
(%) |
range |
|
|
|
|
|
|
|
|
|
|
4 |
63 |
45 |