Intact HCG, free HCG {beta} subunit and HCG {beta} core fragment: longitudinal patterns in urine during early pregnancy

Ruth McChesney1, Allen J. Wilcox2,8, John F. O'Connor3, Clarice R. Weinberg4, Donna D. Baird2, John P. Schlatterer3, D.Robert McConnaughey5, Steven Birken6 and Robert E. Canfield7

1 Brooklyn College, The City University of New York, 3 Irving Center for Clinical Research and 6 Department of Obstetrics and Gynaecology, 7 Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, NY, 2 Epidemiology Branch and 4 Biostatistics Branch, NIEHS and 5 WESTAT, Durham, NC, USA

8 To whom correspondence should be addressed at: Epidemiology Branch MD A3-05, National Institute of Environmental Health Sciences (NIEHS), Durham, NC 27709, USA. Email: wilcox{at}niehs.nih.gov


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Detecting and monitoring early pregnancy depend on the measurement of HCG. Little is known about how production of various forms of HCG may evolve over the earliest weeks of pregnancy, particularly in naturally conceived pregnancies. METHODS: We describe the daily excretion of three urinary HCG analytes during the first 6 weeks post-conception in 37 naturally conceived pregnancies ending in singleton birth. We assayed daily first morning urine samples for intact HCG, free {beta} subunit and core fragment, plus the combined measurement of these HCG forms. We calculated doubling times for each analyte and the inter- and intra-subject day-to-day variation. RESULTS: Intact HCG and the free {beta} subunit were initially the predominant forms of HCG, with the {beta} core fragment emerging as the predominant form in the fifth week after conception. Intact HCG and the free {beta} subunit showed the most day-to-day variability, and were transiently undetectable even 10 days after detection of pregnancy. The most stable estimate of doubling time was provided by the combined measurement of all these forms. CONCLUSIONS: Although intact HCG is usually regarded as the main analyte for detection and monitoring of early pregnancy, it can fluctuate markedly during early pregnancy. This variability could affect pregnancy test results based on early pregnancy urine, and may distort estimates of doubling time. Assays that combine several forms of HCG may be more reliable.

Key words: HCG/longitudinal patterns/pregnancy/urine


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
HCG is the cornerstone of early pregnancy detection and monitoring. The most well characterized form of HCG is the intact heterodimer (intact HCG), composed of an {alpha} and {beta} subunit. However, there are other isoforms in pregnancy including the free HCG {alpha} subunit, the free HCG {beta} subunit (HCG{beta}) and the HCG {beta} core fragment (HCG{beta}cf) (Birken et al., 1988Go; Blithe et al., 1988Go). Microheterogeneity among these forms has also been described. Variants can differ in glycosylation (with hyperglycosylated forms) or sialic acid content, and peptide bond cleavages in the {beta} subunit may give rise to ‘nicked’ molecules (O'Connor et al., 1994Go; Elliott et al., 1997Go; Kovalevskaya et al., 1999Go).

There is a well-documented exponential rise in HCG following implantation (Mishell et al., 1974Go; Lenton et al., 1982Go; Pittaway et al., 1985Go; Fritz and Guo, 1987Go). With few exceptions, HCG measurements from the earliest stages of pregnancy have been derived from patients attending clinics for fertility treatment. Very little information is available on naturally conceived pregnancies, especially during the earliest weeks of pregnancy (Cole et al., 1993Go; O'Connor et al. 1998Go; Mock et al., 2000Go). We know of no data describing the serial daily production of HCG isoforms during the first weeks of pregnancy.

We assayed several forms of HCG to explore the daily patterns of urinary HCG excretion in the first 6 weeks following conception. These samples came from 37 successful natural pregnancies of fertile women who participated in the North Carolina Early Pregnancy Study.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Subjects
The Early Pregnancy Study was conducted in North Carolina, USA from 1982 to 1986. This was a prospective study of 221 women with no known chronic health or fertility problems who were planning to become pregnant. Women were enrolled at the time they discontinued their use of any method of birth control. Women collected daily first morning urine samples, which they stored in their home freezers. Women participated for up to 6 months. If they became clinically pregnant in this time, they continued urine collection for a minimum of 8 weeks after their last menstrual period. This study was approved by the Institutional Review Board of the National Institute of Environmental Health Sciences, and all participants provided informed consent. Details of the study have been reported (Wilcox et al., 1985Go, 1988Go).

We re-contacted women who had conceived around their expected time of delivery to determine the outcome of pregnancy. There were 130 women who delivered live born singleton infants. In order to conduct more extensive analyses on a subset of these 130 women, 50 pregnancies were drawn by random selection. We selected from these the pregnancies with the most complete urine samples on crucial days, providing 37 pregnancies for the present analysis.

Specimen collection
Subjects collected daily first morning urine samples in wide-mouth polypropylene jars with screw-on lids (capacity 30 ml, with no preservatives). Details of transport and storage are provided below. Women collected usable urine samples for an average of 98% of their days in the study. Among the 37 women, no subject was missing more than three consecutive days of urine samples.

Estrogen and progesterone: day of ovulation
Urinary estrone-3-glucuronide (E1G) and pregnanediol-3-glucuronide (PdG) were determined by radioimmunoassay. This information was used to identify the day of ovulation, based on the characteristic changes in the ratio of estrogen and progesterone around the time of ovulation. The use of the steroid ratio as a marker of ovulation was developed and validated by measurements of LH (Baird et al., 1991Go, 1995Go; Dunson et al., 2001Go). This method has been validated subsequently against ultrasound determination of ovulation (Ecochard et al., 2001Go). A day of ovulation was identified for all 37 conception cycles.

HCG: implantation and early pregnancy
All analytes of HCG were identified by immunoradiometric assays (IRMAs). Methods for each specific assay are described below. Urine from a pre-pubescent male donor was used as a diluent for standards and samples to control for matrix effects. The capture and detection antisera are indicated for each HCG analyte assay (with the detection antibody indicated by *). Assay sensitivity (least detectable dose) was defined as 2 SDs higher than the non-specific binding. All assays were carried out at the College of Physicians and Surgeons, Columbia University, New York. Table I provides a summary of antisera used in each HCG IRMA, as well as the cross-reactivities with other major HCG or LH analytes.


View this table:
[in this window]
[in a new window]
 
Table I. Analyses of HCG by immunoradiometric assay (IRMA)

 
Sepharose-IRMA B101-R525*. This is the original assay used to identify HCG in the North Carolina study (Armstrong et al., 1984Go; Canfield et al., 1987Go). The IRMA was based on Sepharose-B101 (conjugated to Sepharose CNBr) capture with detection by the polyclonal antiserum R525 directed against the C-terminal region of HCG{beta}. This combination of antisera primarily detects intact HCG but also cross-reacts with HCG{beta} (~70%). The lowest detectable concentration for this assay was 0.01 ng/ml (0.27 pmol/l) and the inter-assay coefficient of variation (CV) was 15% (Wilcox et al., 1988Go). The day of first rise in HCG as measured by this assay was considered a marker for time of implantation (Wilcox et al., 1999Go). The day of implantation was identified for 36 of the 37 subjects.

Microtitre IRMA. Details of these methods have been reported previously (O'Connor et al., 1988Go; Krichevsky et al., 1991Go). In summary, a 96-well microtitre plate format was employed using holders with Immulon-2 microtitre strips (Fisher Scientific, Springfield, NJ). The capture antibody solution (0.2 ml in binding buffer, 0.2 mol/l NaHCO3, pH 9.5) was added to the wells and incubated overnight at 4 °C. After removal of the antibody solution, blocking solution [1% bovine serum albumin (BSA) containing 0.01% NaN3] was then added for 3 h at room temperature (plates can be stored at 4 °C for up to 4 weeks at this stage). The BSA was removed and the wells washed with deionized water. Matrix effects were controlled for by the use of normal male urine as a diluent for standards and blanks. All urines and standards were adjusted to pH 8.0 with 1 mol/lM Tris buffer, pH 9.0, and 200 µl of a range of standards or 200 µl of unconcentrated or dilutions of urine were then added to the wells. After incubation for 24 h at room temperature, the wells were washed with deionized water, and 200 µl of iodinated detection antibody (40 000 c.p.m./tube) in assay buffer (phosphate-buffered saline, 0.01 mol/l EDTA, 0.01 mol/l NaN3, 0.1% bovine {gamma}-globulin) was added. Incubation was carried out for 24 h at room temperature. Following aspiration of the unbound trace, the wells were washed five times with deionized water. Radioactivity was determined in a Packard gamma counter. Values were interpolated from a smoothed spline transformation of the standards. For computational purposes, non-detectable HCG analytes were assigned a value defined as 50% of the least detectable dose.

Combination B109/B204-B108*. This assay was constructed to detect three major HCG metabolites: intact HCG (with or without the presence of the C-terminal peptide), HCG{beta} and HCG{beta}cf (O'Connor et al, 1988Go). Although the B109 capture antibody does not detect the ‘nicked’ form of intact HCG, some degree of detection is afforded by B204, since nicking apparently alters the architecture of the intact molecule to expose HCG{beta} and HCG{beta}cf epitopes. The standard was CR127, with a range of 2.6–341 pmol/l. Detection limits for the assay were 1–2 pmol/l. The intra-assay CV was 5.7%. Inter-assay CVs for 6.8, 68 and 273 pmol/l concentrations were 15.0, 11.9 and 10.2%, respectively.

Intact HCG heterodimer B109-B108*. This assay detected the intact non-nicked heterodimeric form of HCG. Standard curves covered the range of 2.6–341 pmol/l and used CR127 as standard. Detection limits were 1.1–1.6 pmol/l. The intra-assay CV was 6.2% and inter-assay CVs for 6.8, 68 and 273 pmol/l were 18.5, 11.6 and 9.8%, respectively. The assay has <1% cross-reactivity with HCG{beta}, HCG{beta}cf, nicked HCG (HCGn), nicked free HCG{beta} (HCG{beta}n), intact human LH (hLH), free hLH {beta} subunit (hLH{beta}) and hLH {beta} core fragment (hLH{beta}cf) purified from the pituitary (Birken et al., 1993Go).

Free HCG {beta} subunit B201-CTP104*. This assay does not differentiate between nicked or non-nicked HCG free {beta} subunit. Detection limits were 10–27 pmol/l. The intra-assay CV was 5.5% and the inter-assay CVs for 54, 230 and 653 pmol/l were 16.8, 11.7 and 11.2%, respectively. The cross-reactivity is 1% with intact HCG and <1% with HCG{beta}cf, hLH, hLH{beta} and hLH{beta}cf (pituitary).

HCG {beta} core fragment B210-B108*. Detection limits were 1.1–2.2 pmol/l. The intra-assay CV was 5.3% and the inter-assay CVs for 7.5, 26 and 79 pmol/l were 11.0, 9.3 and 13.7%, respectively. There is 2% cross-reactivity with hLH{beta}cf (pituitary) and <1% with intact HCG HCG{beta}, HCGn, HCG{beta}n, hLH and hLH{beta}.

Handling and preparation of samples for assay
Women placed first morning urine specimens directly into home freezers. Study personnel visited their homes regularly (usually weekly) to pick up frozen specimens. Specimens were placed in insulated containers with dry ice and transported to a central freezer, where they were maintained and monitored at –20 °C. Specimens were shipped to collaborating laboratories by overnight freight in insulated containers with dry ice. All laboratory storage was in monitored freezers at –20 or –80 °C.

At 6–24 months following collection, specimens were thawed for initial HCG immunoradiometric (R525) assay and refrozen (1983–1987). In 1986–1989, urine specimens were thawed for steroid hormone assay, at which time they were also aliquoted into smaller containers.

When specimens were 4–10 years old (1990–1992), they were thawed for a third time and assayed for the remaining HCG analytes described above. All dilutions for those assays were prepared at once for each sample, and used for the entire panel of analyte assays.

HCG stability
We assessed the stability of the intact HCG, HCG{beta} and HCG{beta}cf by subjecting purified analytes in urine to 40 cycles of freezing and thawing and to storage at 4 °C for 4 weeks. The three HCG analytes were assayed at seven stages during the freeze–thaw process, and at seven points in time. There was no loss of immunodetection at any stage of freezing and thawing, or at any point in time with storage at 4 °C.

Creatinine
Creatinine measurements were performed for daily samples through the third week after ovulation, and every third day subsequently.

Data analysis
Data analyses were carried out using SAS (Cary, NC). HCG data were log transformed, and all means were calculated as geometric means. The doubling time for HCG rise (Batzer et al., 1981Go) was estimated on the basis of HCG measurements on days 14–21 after ovulation, and days 5–12 after estimated implantation.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The 37 women in this analysis were between 23 and 40 years of age, with a mean age of 30 (Table II). Thirty percent were nulliparous at enrolment. Overall, these women were representative of the entire study group in age, reproductive history and other characteristics (Wilcox et al., 1988Go). Their first day of HCG rise ranged from 6 to 12 days after ovulation (median day 9), the same as for the whole sample (Wilcox et al., 1999Go).


View this table:
[in this window]
[in a new window]
 
Table II. Characteristics of subjects at enrolment in the study (n=37)

 
In the original study, HCG detection had been limited to the assay of intact HCG by the Sepharose-IRMA, and included only the first 7 days following implantation. The present study presents three new assays for individual HCG forms plus a combination assay that includes the three forms. Also, we extended our urine analyses to include up to 6 weeks of daily samples after implantation.

Data for the various HCG forms are shown in Figure 1. The reference date (day 0) is the day of ovulation, which is almost certainly also the day of conception (Wilcox et al., 1999Go). The HCG values starting from this date are shown on a log scale, which accommodates the exponential rise of HCG. (We provide results from the original IRMA assay for the first 7 days of pregnancy, as a reference.)



View larger version (20K):
[in this window]
[in a new window]
 
Figure 1. Geometric means for daily concentrations of urinary HCG analytes (pmol/l), relative to day of ovulation (day 0).

 
All forms of HCG showed an exponential rise after the first day of detection, with a general pattern of decelerating rates of rise by the fifth week. However, within this general pattern, there were striking differences in the rates of rise. Although there was some immunodetection by the HCG{beta}cf assay just after ovulation, the HCG{beta}cf was lowest in concentration for the first few weeks and continued to rise steadily as the others slowed. By the fifth week after conception, the HCG{beta}cf had emerged as the predominant form of HCG. The extent of differences among these HCG forms may not be fully appreciated in Figure 1 due to the log scale. For example, on day 14, mean concentration (in pmol/l) was 8 for HCG{beta}cf, 21 for intact HCG and 297 for HCG{beta} (Table III). On day 35, the respective values were 65 000, 16 000 and 22 000 pmol/l.


View this table:
[in this window]
[in a new window]
 
Table III. Measures of variation in urinary HCG analytes: coefficients of variation (CVs) and geometric means±SEM are shown

 
The forms of HCG also differed in their day-to-day variability (Figure 1). The combination HCG assay showed a relatively smooth progression overall, as did the HCG{beta}cf. In contrast, the rise of intact HCG and HCG{beta} was more irregular. We examined these patterns more closely by looking at data for individual women. As in the summary data, the smoothest patterns of rise were found with the combination assay and the HCG{beta}cf, while intact HCG and HCG{beta} could be highly variable from day to day. Data for three women are shown in Figure 2. Even in the second week after implantation, the HCG{beta} and intact HCG sporadically fell below the detection level. For woman A, intact HCG dropped to undetectable levels on days 10 and 11 after the start of the HCG rise. For woman B, HCG{beta} was undetectable on a day 3 weeks after the start of the rise.



View larger version (30K):
[in this window]
[in a new window]
 
Figure 2. Daily profiles of HCG analytes (pmol/l) and creatinine (g/l) during early pregnancy for 3 women. Day 0 represents the day of ovulation.

 
There are several possible explanations for this extreme variability of the HCG{beta} and intact HCG. One is variation in urinary concentration. This seems unlikely, given that within a woman, wide excursions occurred on different days for different analytes. However, we explored this using creatinine concentration data available for about half of the urine samples. The creatinine data are provided in the bottom panel of Figure 2 (using the same y-axis scale as for the HCG assays). The variability of the HCG{beta} and intact HCG is far greater than the variations seen in creatinine.

Another possibility is assay variability. However, the inter-assay CV was similar for all the microtitre assays; it therefore seems unlikely that fluctuations in two of the assays could be due to batch differences. The variation in intact HCG could not be explained by the standard used, because the combination assay and intact HCG assay each employed the same purified intact HCG preparation. Adjustment with buffer prior to assay controlled for urinary pH extremes.

A more likely explanation is assay error. We therefore reanalysed samples from 12 pregnancies using aliquots of original samples that had not been thawed previously for assay. We re-assayed 3–16 days per woman. In choosing days for repeat analyses, we specifically included days with extreme changes. In every case, the repeat assays showed exactly the same extreme fluctuations in HCG concentration.

Finding no basis to dismiss the day-to-day fluctuations within women, we looked for ways to quantify them. First, we determined the variance of HCG concentrations within the window from 18 to 28 days after fertilization. This time is well beyond the early post-implantation phase, and therefore a time when HCG measures would presumably be more dependable. Figure 3 shows the distribution of day-to-day variances (log scale) among the 37 women. Intact HCG showed by far the largest range of day-to-day changes, with the variance ranging widely across women.



View larger version (33K):
[in this window]
[in a new window]
 
Figure 3. Day-to-day variance among 37 women for each HCG analyte, days 18–28 post-ovulation.

 
We looked at CVs as yet another way to summarize variability within HCG analytes. These data are shown in Table III. The first analysis selected all HCG values on day 14 after ovulation. The highest CV was found for intact HCG. Adjustment for creatinine modestly reduced the CV but did not change the pattern.

One artifact that can contribute to a general variation of HCG across pregnancies is the individual differences in time from ovulation to implantation (ranging from 6 to 12 days; Wilcox et al., 1999Go). We removed this variation by re-setting the time scale for each pregnancy to start on the day of implantation. We then selected HCG values on day 5 after implantation, which corresponds most closely to day 14 after ovulation (mean time from ovulation to implantation is 9 days). As expected, the CVs were reduced by this adjustment, but intact HCG still had the highest CV. Further adjustment for creatinine did not change this pattern of strikingly high variability for HCG.

We assessed the rates of rise for each HCG analyte using doubling time. Table IV shows the doubling times during the window from 14 to 21 days after ovulation (~5–12 days after implantation). The HCG{beta}cf had a faster doubling time than either HCG{beta} or intact HCG. The intra-woman fluctuations in intact HCG (and to a lesser extent, in HCG{beta}) were expressed as a broader range of doubling times for these analytes.


View this table:
[in this window]
[in a new window]
 
Table IV. Doubling times (in days) for each HCG analyte calculated for time intervals from days 14 to 21 post-ovulation, or days 5 to 12 after implantation

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The progression of HCG during early pregnancy can be described only with frequent sampling of biological specimens. Our study of 37 naturally conceived pregnancies ending in live birth constitutes a relatively large and complete set of data for this purpose. Our results are based on collection of urine samples, which is a practical necessity when measuring HCG daily in healthy women. Even though blood may generally be preferred over urine for endocrinological assays, urine assays have proved to perform well for HCG measurements. Urinary concentrations of intact HCG are similar to serum concentrations (Wehmann and Nisula, 1981Go; Norman et al., 1987Go). We observe that doubling times for urinary HCG in early pregnancy are similar to those determined for serum measurements (Pittaway et al., 1985Go). In some circumstances, urine measures of HCG may even be superior to serum measures (Olsen et al., 2001Go). First morning voids of urine provide an integrated measure over time, thus smoothing possible pulsatile secretion of HCG by the trophoblast (Diaz-Cueto et al., 1994Go). The fact that commercial pregnancy tests and clinical kits rely on measurements of urinary HCG is yet another reason to be interested in the patterns of HCG excretion in urine during early pregnancy.

Our measurements of HCG are based on well-characterized recognition sites for capture and detection antisera, as well as appropriate purified standards in each analyte assay (O'Connor et al., 1988Go, 1994Go). Each analyte being captured probably comprises a variety of isoforms. One example is hyperglycosylated HCG (H-HCG). This family of HCG variants has been the subject of much recent study (Weinans et al., 2000Go; Kovalevskaya et al., 2002Go), although the variants are not yet fully characterized. For example, it is not yet known how many H-HCG isoforms may be present in early pregnancy.

The only forms of H-HCG that can be studied at present are those identified by the B-152 antibody. These forms are hyperglycosylated in the C-terminal region of the {beta} subunit. Our capture antibody (B-109) does not bind near the C-terminal region, and therefore should not be affected by the structure of the C-terminal peptide. Based on epitope maps of HCG, this H-HCG should be fully captured by our assays for intact HCG and HCG{beta} (Birken et al., 2003Go).

Final proof of this awaits the development of a true standard for H-HCG in pregnancy. What is currently regarded as ‘standard’ H-HCG has been derived from choriocarcinoma, and has a different molecular weight from the form found in pregnancy (Kovalevskaya et al., 2002Go). Also, the H-HCG derived from cancer cells is more likely to be ‘nicked’ HCG (another isoform variant) than the HCG in early pregnancy (Kovalevskaya et al., 2002Go). Once valid standards have been developed for H-HCG and its variants, more detailed studies of their patterns in early pregnancy will be possible.

Our data describe the daily progression of several forms of HCG during the first 5 weeks after conception. In the earliest weeks of pregnancy, the HCG{beta}cf has the lowest concentration of any of the measured HCG analytes. When HCG{beta}cf has been generally regarded as the predominant form of urinary HCG in pregnancy (Kato and Braunstein, 1988Go; de Medeiros et al., 1992Go), this conclusion is based on samples collected later in pregnancy. Our data clearly show that the HCG{beta}cf emerges as the dominant form only during the fifth week after conception. Conversely, HCG{beta} was relatively common in the first 3 weeks after implantation, and less common thereafter.

Given that the urine samples in our study were stored and handled uniformly, it is unlikely that the relative changes between analytes that we observe during the first 6 weeks of early pregnancy could be due to degradation, disassociation or other changes attributable to storage. In particular, the relative gain in HCG{beta} over HCG{beta}cf during the first few weeks of pregnancy is not easily explained by changes due to storage.

Another unexpected finding was the high day-to-day variability of intact HCG in urine, both across women and within individual women. This variability was not due to physiological differences in urine concentration, and was confirmed using duplicate urine samples that had not been thawed previously. These abrupt changes cannot be explained by our current understanding of metabolism and urinary clearance (Wehman and Nisula, 1981Go). It is theoretically possible that an interfering factor in urine could disturb the binding to the specific capture antibody. No exogenous chemicals had been added to the specimens as preservatives, so interfering substances would have to be endogenous or the result of contamination of the urine sample during voiding or storage. These explanations seem unlikely given the sporadic nature of the fluctuations, and the fact that these fluctuations tended to occur on different days for different analytes for the same woman.

Day-to-day changes in a woman's diet, personal habits or exposures might contribute to variability in urinary excretion of specific HCG analytes, although few examples are known. Levels of HCG (albeit later in pregnancy) have been reported to be decreased by cigarette smoking (Bernstein et al., 1989Go; Bremme et al., 1990Go). Only two of these 37 women reported smoking. We lack the detailed dietary or other daily information that would be needed to pursue such possibilities.

Stability of these analytes over time must be considered in assessing our results, given that the urine specimens were 4–10 years old at the time of assay. The analytes we measured are among the most stable analytes of HCG. In a previous test of stability, we found that the concentration of intact HCG actually increased slightly over time, probably due to sublimation of water (Wilcox et al., 1985Go). Direct assessment of the stability of intact HCG, HCG{beta} and the HCG{beta}cf showed no evidence of degradation under more stressful experimental conditions than experienced by our samples.

If the unpredictable changes in urinary concentrations of intact HCG over successive days are a general occurrence in early pregnancy, there are clinical implications. Intact HCG has usually been regarded as the most bioactive of the several forms of HCG. More importantly, it provides the basis for many urinary assays for detection of pregnancy (Chard, 1992Go; Cole et al., 1993Go; Butler et al., 2001Go). To the degree that a pregnancy test relies solely on the detection of HCG, a single test could be falsely negative even a week or more after implantation (see ‘intact HCG’, Figure 2). This may be relevant to the performance of commercial urine-based pregnancy test kits (of which nearly 20 million are sold annually in the USA).

Similarly, clinical measurements of HCG rate of rise based on urine assays might be distorted dramatically by the variability of intact HCG, especially if the rate of rise is determined on the basis of only two tests (as is often the case). These limitations of intact HCG could be compensated by expanding assays to detect additional forms of HCG. In our study, the combination assay appears to be extremely reliable and sensitive (Figure 1). [Our earlier Sepharose-IRMA was based on intact HCG and no doubt benefited from its cross-reactivity with HCG{beta} (Wilcox et al.., 1988Go).]

There has been little discussion of sources of variation in urine HCG analyses (Lopata et al., 1982Go; Cole et al., 1993Go; Mishalani et al., 1994Go; Butler et al., 2001Go). It is possible that these fluctuations have biological significance, and that the physiological roles of intact HCG and its several forms may be more complex than previously suspected. For example, there could be paracrine or autocrine functions of HCG that evolve over the course of early pregnancy. If such mechanisms are operating, they are unlikely to be discovered without frequent sampling during the first weeks of pregnancy—ideally in serum as well as in urine, and using a battery of assays that describe multiple forms of HCG.

In conclusion, our data show a gradually changing pattern of HCG forms excreted during early pregnancy, and reveal unexplained fluctuations in the day-to-day excretion of intact HCG (and to a lesser extent HCG{beta}). Variations over time among women and within women contribute to wide ranges of concentrations for intact HCG during the early weeks of pregnancy. The biological mechanisms of these variations and their functional significance are unknown at present. However, pregnancy tests that depend on the detection of urinary HCG may be more reliable and valid if they capture several HCG forms. This strategy is likely to provide a more sensitive assay, as well as one less subject to the apparent vagaries of concentration found with the intact HCG in early pregnancy.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
We thank Joy Pierce for the management of the field study and urine collections, Paul Musey and Delwood Collins for performing the steroid and creatinine assays, and Freya Kamel for providing helpful comments on an earlier draft of the manuscript.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Armstrong EG, Ehrlich PH, Birken S, Schlatterer JP, Siris E, Hembree WC and Canfield RE (1984) Use of a highly sensitive and specific immunoradiometric assay for detection of human chorionic gonadotropin in urine of normal, nonpregnant, and pregnant individuals. J Clin Endocrinol Metab 59, 867–874.[Abstract]

Baird DD, Weinberg CR, Wilcox AJ, McConnaughey DR and Musey PI (1991) Using the ratio of urinary oestrogen and progesterone metabolites to estimate day of ovulation. Stat Med 10, 255–266.[ISI][Medline]

Baird DD, McConnaughey DR, Weinberg CR, Musey PI, Collins DC, Kesner JS, Knecht EA and Wilcox AJ (1995) Application of a method for estimating day of ovulation using urinary estrogen and progesterone metabolites. Epidemiology 6, 547–550.[ISI][Medline]

Batzer FR, Schlaff S, Goldfarb AF and Corson SL (1981) Serial {beta}-subunit human chorionic gonadotropin doubling time as a prognosticator of pregnancy outcome in an infertile population. Fertil Steril 35, 307–312.[ISI][Medline]

Bernstein L, Pike ML, Lobo RA, Depue RT, Ross RK and Henderson BE (1989) Cigarette smoking in pregnancy results in marked decrease in maternal hCG and oestradiol levels. Br J Obstet Gynaecol 96, 92–96.[ISI][Medline]

Birken S, Armstrong EG, Kolks MAG, Cole LA, Agosto GM, Krichevsky A, Vaitukaitis JL and Canfield RE (1988) Structure of the human chorionic gonadotropin {beta}-subunit fragment from pregnancy urine. Endocrinology 123, 572–583.[Abstract]

Birken S, Chen Y, Gawinowicz MA, Agosto GM, Canfield RE and Hartree AS (1993) Structure and significance of human luteinizing hormone-{beta} core fragment purified from human pituitary extracts. Endocrinology 133, 985–989.[Abstract]

Birken S, Yershova O, Myers RV, Bernard MP and Moyle W (2003) Analysis of human choriogonadotropin core 2 o-glycan isoforms. Mol Cell Endocrinol 204, 21–30.[CrossRef][ISI][Medline]

Blithe DL, Akar AH, Wehmann RE and Nisula BC (1988) Purification of {beta}-core fragment from pregnancy urine and demonstration that its carbohydrate moieties differ from those of native human chorionic gonadotropin-{beta}. Endocrinology 122, 173–180.[Abstract]

Bremme K, Lagerstrom M, Andersson O, Johansson S and Eneroth P (1990) Influences of maternal smoking and fetal sex on maternal serum oestriol, prolactin, hCG, and hPL levels. Arch Gynecol Obstet 247, 95–103.[ISI][Medline]

Butler SA, Khanlian SS and Cole LA (2001) Detection of early pregnancy forms of human chorionic gonadotropin by home pregnancy test devices. Clin Chem 47, 2131–2136.[Abstract/Free Full Text]

Canfield RE, O'Connor JF, Birken S, Krichevsky A and Wilcox AJ (1987) Development of an assay for a biomarker of pregnancy and early fetal loss. Environ Health Perspect 74, 57–66.[ISI][Medline]

Chard T (1992) Pregnancy tests: a review. Hum Reprod 7, 701–719.[Abstract]

Cole LA, Seifer DB, Kardana A and Braunstein GD (1993) Selecting human chorionic gonadotropin immunoassays: consideration of cross-reacting molecules in first-trimester pregnancy and urine. Am J Obstet Gynecol 163, 1580–1586.

de Medeiros SF, Amato F, Matthews CD and Norman RJ (1992) Urinary concentrations of beta core fragment of hCG throughout pregnancy. Obstet Gynecol 80, 223–230.[Abstract]

Diaz-Cueto L, Mendez JP, Barrios-de-Tomasi J, Lee J-Y, Wide L, Veldhuis JD and Ulloa-Aguirre A (1994) Amplitude regulation of episodic release, in vitro biological to immunological ratio, and median charge of human chorionic gonadotropin in pregnancy. J Clin Endocrinol Metab 87, 890–897.[CrossRef]

Dunson DB, Weinberg CR, Baird DD, Kesner JS and Wilcox AJ (2001) Assessing fertility using several markers of ovulation. Stat Med 20, 965–978.[CrossRef][ISI][Medline]

Echochard R, Boehringer H, Rabilloud M and Marret H (2001) Chronological aspects of ultrasonic, hormonal, and other indirect indices of ovulation. Br J Obstet Gynaecol 108, 822–829.[CrossRef][ISI]

Elliott MM, Kardana A, Lustbader JW and Cole LA (1997) Carbohydrate and peptide structure of the alpha and beta subunits of chorionic gonadotropin (hCG): characteristics and variants in 32 subunit preparations from normal and aberrant pregnancy and choriocarcinoma. Endocrine 7, 15–32.[ISI][Medline]

Fritz MA and Guo S (1987) Doubling time of human chorionic gonadotropin (hCG) in early normal pregnancy: relationship to hCG concentration and gestational age. Fertil Steril 47, 584–589.[ISI][Medline]

Kato Y and Braunstein GD (1988) {beta}-Core fragment is a major form of immunoreactive urinary chorionic gonadotropin in human pregnancy. J Clin Endocrinol Metab 66, 1197–1201.[Abstract]

Kovalevskaya G, Birken S, Kakuma T, Schlatterer J and O'Connor JF (1999) Evaluation of nicked human chorionic gonadotropin content in clinical specimens by a specific immunometric assay. Clin Chem 45, 68–77.[Abstract/Free Full Text]

Kovalevskaya G, Birken S, Kakuma T, Ozaki N, Sauer M, Lindheim S, Cohen M, Kelly A, Schlatterer J and O'Connor JF (2002) Differential expression of human chorionic gonadotropin (hCG) glycosylation isoforms in failing and continuing pregnancies: preliminary characterization of the hyperglycosylated hCG epitope. J Endocrinol 172, 497–506.[Abstract/Free Full Text]

Krichevsky A, Birken S, O'Connor J, Bikel K, Schlatterer J, Chen Y, Agosto G and Canfield R (1991) Development and characterization of a new, highly specific antibody to the human chorionic gonadotropin {beta}-subunit. Endocrinology 128, 1255–1264.[Abstract]

Lenton EA, Neal LM and Sulaiman R (1982) Plasma concentrations of human chorionic gonadotropin from the time of implantation until the second week of pregnancy. Fertil Steril 37, 773–778.[ISI][Medline]

Lopata A, Martin M, Oliva K and Johnston I (1982) Embryonic development and blastocyst implantation following in vitro fertilization and embryo transfer. Fertil Steril 38, 682–687.[ISI][Medline]

Mishalani SH, Seliktar J and Braunstein GD (1994) Four rapid serum-urine combination assays of choriogonadotropin (hCG) compared and assessed for their utility in quantitative determinations of hCG. Clin Chem 40, 1944–1949.[Abstract/Free Full Text]

Mishell DR, Nakamura RM, Barberia JM and Thorneycroft IH (1974) Initial detection of human chorionic gonadotropin in serum in normal human gestation. Am J Obstet Gynecol 118, 990–991.[ISI][Medline]

Mock P, Kovalevskaya G, O'Connor JF and Campana A (2000) Choriocarcinoma-like human chorionic gonadotrophin (HCG) and HCG bioactivity during the first trimester of pregnancy. Hum Reprod 15, 2209–2214.[Abstract/Free Full Text]

Norman RJ, Menabawey M, Lowings C, Buck RH and Chard T (1987) Relationship between blood and urine concentrations of intact human chorionic gonadotropin and its free subunits in early pregnancy. Obstet Gynecol 69, 590–593.[Abstract]

O'Connor JF, Schlatterer JP, Birken S, Krichevsky A, Armstrong EG, McMahon D and Canfield RE (1988) Development of highly sensitive immunoassays to measure human chorionic gonadotropin, its {beta}-subunit, and {beta}-core fragment in the urine: application to malignancies. Cancer Res 48, 1361–1366.[Abstract]

O'Connor JF, Birken S, Lustbader JW, Krichevsky A, Chen Y and Canfield RE (1994) Recent advances in the chemistry and immunochemistry of human chorionic gonadotropin: impact on clinical measurements. Endocr Rev 15, 650–683.[ISI][Medline]

O'Connor JF, Ellish N, Kakuma T, Schlatterer J and Kovalevskaya G (1998) Differential urinary gonadotrophin profiles in early pregnancy and early pregnancy loss. Prenat Diagn 18, 1232–1240.[CrossRef][ISI][Medline]

Olsen TG, Hubert PR and Nycum LR (2001) Falsely elevated human chorionic gonadotropin leading to unnecessary therapy. Obstet Gynecol 98, 843–845.[Abstract/Free Full Text]

Pittaway DE, Reish RL and Wentz AC (1985) Doubling times of human chorionic gonadotropin increase in early viable intrauterine pregnancies. Am J Obstet Gynecol 152, 299–302.[ISI][Medline]

Wehmann RE and Nisula BC (1981) Metabolic and renal clearance rates of purified human chorionic gonadotropin. J Clin Invest 68, 184–194.[ISI][Medline]

Weinans MJ, Butler SA, Mantingh A and Cole LA (2000) Urinary hyperglycosylated hCG in first trimester screening for chromosomal abnormalities. Prenat Diagn 20, 976–978.[CrossRef][ISI][Medline]

Wilcox AJ, Weinberg CR, Wehmann RE, Armstrong EG, Canfield RE and Nisula BC (1985) Measuring early pregnancy loss: laboratory and field methods. Fertil Steril 44, 366–374.[ISI][Medline]

Wilcox AJ, Weinberg CR, O'Connor JF, Baird DD, Schlatterer JP, Canfield RE, Armstrong EG and Nisula BC (1988) Incidence of early loss of pregnancy. N Engl J Med 319, 189–194.[Abstract]

Wilcox AJ, Baird DD and Weinberg CR (1999) Time of implantation of the conceptus and loss of pregnancy. N Engl J Med 340, 1796–1799.[Abstract/Free Full Text]

Submitted on June 23, 2003; resubmitted on December 23, 2003; accepted on December 2, 2004.





This Article
Abstract
Full Text (PDF )
All Versions of this Article:
20/4/928    most recent
deh702v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by McChesney, R.
Articles by Canfield, R. E.
PubMed
PubMed Citation
Articles by McChesney, R.
Articles by Canfield, R. E.