Clinical Chemistry 49:12
2045–2049 (2003)
Endocrinology and
Metabolism
Human Chorionic Gonadotropin Isoforms in the
Diagnosis of Ectopic Pregnancy
Paola T.A. Borrelli,1,2 Stephen A. Butler,1,3 Suzanne M. Docherty,1,4 Edyta M. Staite,1
Antonio L. Borrelli,2 and Ray K. Iles1,4*
Background: Early diagnosis of ectopic pregnancy uses
ultrasound with serial measurements of total human
chorionic gonadotropin (hCG). The objective of this
study was to explore the possibility that an isolated
measurement of hCG isoforms/subunits rather than
total hCG could be used as a single test for ectopic
pregnancy.
Methods: Total and intact hCG, free hCG - and ␣-subunits (hCG and -␣), and hCG -core fragment were
measured by RIA and IRMA in the serum and urine of
76 women presenting at outpatient emergency departments with a positive pregnancy test, lower abdominal
pain, and/or vaginal bleeding. Final diagnoses were
based on outcomes of pregnancies and tissue histology.
Results: Twenty-seven of the 76 women were subsequently diagnosed with viable pregnancies, 37 with
spontaneous miscarriage, and 12 with ectopic pregnancy. Concentrations of all forms of hCG were lower in
cases of ectopic pregnancy and spontaneous miscarriage
than in viable pregnancies. Serum samples gave better
results than urine samples. The free hCG isoform (P
<0.0001) had 100% sensitivity at a specificity of 79% at a
281 pmol/L (6.5 g/L) cutoff. Total hCG (P ⴝ 0.005) had
comparable ROC characteristics with a 100% sensitivity
and 68% specificity at a cutoff value of 1053 pmol/L (375
IU/L). Neither hCG (P ⴝ 0.7) nor total hCG (P ⴝ 0.4)
1
Williamson Laboratory, Department of Obstetrics and Gynaecology,
Barts and The London Queen Mary School of Medicine, London EC1A 7BE,
UK.
2
Centro di Diagnosi Prenatale, Dipartimento di Ginecologia, Ostetricia e
Neonatologia, Seconda Università degli Studi di Napoli, Naples, Italy.
3
Diagnostica Medica, Centro di Fisiopatologia della Reproduzione, Mercogliano (AV), Italy.
4
School of Health and Social Science, Middlesex University, Trent Park,
Bramley Road, London, UK.
*Address correspondence to this author at: Williamson Laboratory, Department of Obstetrics and Gynaecology, East Wing, St. Bartholomew’s
Hospital, West Smithfield, London EC1A 7BE, UK. Fax 44-0207-601-7050;
e-mail r.k.iles@qmul.ac.uk.
Received May 16, 2003; accepted September 17, 2003.
DOI: 10.1373/clinchem.2003.022095
could distinguish ectopic pregnancies from spontaneous miscarriage.
Conclusion: Measurement of serum free hCG at the
time of presentation can identify women with a high
probability of ectopic pregnancy who may benefit from
closer surveillance, reducing the risk of tubal rupture.
© 2003 American Association for Clinical Chemistry
Ectopic pregnancy is the major cause of morbidity and
mortality in women of reproductive age, accounting for
⬃9% of all deaths associated with first-trimester pregnancies and affecting ⬃2% of all pregnancies (1, 2 ). Since the
introduction of laparoscopy in the late 1960s (3 ), considerable effort has been made to define more sensitive and
specific tests for its early diagnosis because ⬎80% of
ectopic pregnancies had been diagnosed after tubal rupture.
Several biochemical and imaging markers have been
studied as single or combined (algorithm) tests for occult
ectopic pregnancy (4 – 6 ). The most common of these
involves the combination of transvaginal ultrasound and
serial measurements of human chorionic gonadotropin
(hCG)5 (7, 8 ). The rate of increase of serum hCG indicates
the presence or absence of viable uterine pregnancies:
hCG doubles every 1.5 days up to 5 weeks after the last
menstrual period, and then every 3.5 days from the 7th
week (or when hCG is ⬎10 000 IU/L) (9 ). Although hCG
concentrations do not increase at this rate in most ectopic
pregnancies, they do in approximately one-third of cases,
and anomalies in the dynamics of the hCG increase are
also related to spontaneous miscarriage (10 ). Despite
these problems, the detection of hCG remains an indispensable tool in the diagnosis of ectopic pregnancy.
In women with inconclusive ultrasonographic findings, a “discriminatory serum hCG zone” has been used,
where a diagnosis of ectopic pregnancy is most likely in
5
Nonstandard abbreviations: hCG, human chorionic gonadotropin; hCG
and hCG␣, hCG - and ␣-subunits, respectively; and hCGcf, hCG -core
fragment.
2045
2046
Borrelli et al.: hCG Isoforms and Ectopic Pregnancy
the absence of demonstrable intrauterine pregnancy but
with an adnexal mass or free fluid in the pouch of
Douglas and serum hCG ⬎6500 IU/L (11 ). With the
advent of transvaginal ultrasound, the hCG cutoff has
decreased to 1000 –2000 IU/L (12 ). Although the introduction of this hCG discriminatory zone appears effective, a particular hCG concentration has yet to be defined
as a good diagnostic indicator for ectopic pregnancy (13 ).
hCG is a glycoprotein heterodimer (14, 15 ). Only intact
hCG, i.e., the ␣,  heterodimer, has biological activity, but
other forms of hCG can be found in the placenta, urine,
and serum (16 ). hCG isoforms and subunits are generally
found as metabolic breakdown products of the intact
hormone, but occasionally free subunits can be expressed
and released directly into the circulation. The most commonly assayed hCG fragments are the free - and ␣-subunits of hCG (hCG and hCG␣, respectively) and the
terminal degradation product of hCG, hCG -core fragment (hCGcf), which is found in urine (17 ). hCG assays
currently used in the diagnosis of ectopic pregnancy
measure so-called total hCG, i.e., intact hCG plus free
hCG.
Diagnostic algorithms for ectopic pregnancies that require serial measurements of hCG over several days
necessitate delays in diagnosis because hCG test are often
done on an outpatient basis, and the risk of tubal rupture
is increased. The aim of our study was to explore whether
measurement of the concentrations of a specific hCG
isoform and/or its subunits in serum and/or urine at a
single point in time could be used to test for ectopic
pregnancy.
Materials and Methods
women and samples
Between September 1998 and September 2000, matched
urine and serum samples were collected at the Homerton
and Whipps Cross Hospitals, London, emergency outpatient departments. Samples were collected from 97
women who presented at the Accident and Emergency
Department with abdominal pain, vaginal bleeding, and a
previous positive hCG test. Twenty-one were excluded
for having no or insufficient matched urine or serum
sample for all tests. Of the remaining 76 women, 27 (36%)
were subsequently diagnosed with viable pregnancies, 37
(48%) with spontaneous miscarriage (11 missed, 6 complete, and 20 incomplete), and 12 (16%) with ectopic
pregnancies. The 12 cases of ectopic pregnancy were
confirmed by surgical intervention and histology.
Serum and urine samples were frozen at ⫺20 °C and
then assayed at the Williamson Laboratory at St. Bartholomew’s Hospital. Samples were free of contamination
and contained no hemolyzed blood. Sample concentrations were determined over several assays when run
undiluted and at multiple dilutions of up to 1 in 10 000 (in
phosphate-buffered saline containing 100 mL/L horse
serum). The CV for all assays were ⱕ5% for the midrange
quality control. At the extremes of the assay range, results
for quality-control samples varied up to 20%. However,
sample results were recorded as the mean analyte concentration from multiple, parallel dilution, determinations.
A laboratory technician and a junior faculty member of
staff in laboratory medicine conducted the index tests.
Data interpretation and quality-control audits were conducted by the Laboratory Director. Clinical data were
collected by a House officer blind to the index test results
and audited by a specialist registrar in Obstetrics and
Gynecology. The North East Thames Region Health Authority Ethic Committee approved the study.
hormone assays
In accordance with IFCC recommendations (18 ), we have
calibrated all assays in terms of picomolar concentration
of the immunoreactive analyte [for a review, see Iles and
Chard (16 )]. For discussion purposes, concentrations in
the commonly used units for the hCG analyte have been
added in parentheses where appropriate.
Total hCG was measured by an in-house RIA with
sheep anti-hCG antiserum (S424) that binds both free
hCG and the intact hormone. Standards were obtained
from the National Institute of Biological Standards (Potters Bar, UK). Free hCG, labeled with 125I by the chloramine-T method, was used as a tracer (NIH preparation
CR123). The detection limit of this assay is 42 pmol/L (15
IU/L) (19 ).
Intact hCG was measured with an in-house IRMA with
a polyclonal anti-hCG␣ capture antibody conjugated to
1,1⬘-carbonyldiimidazole-activated cellulose and a 125Iradiolabeled monoclonal antibody (1/07; Quantum Bioscience) to epitopes on the hCG C-terminal peptide. The
calibration curve ranged from 2.6 to 813 pmol/L (0.93–289
IU/L) (19 ).
Free hCG␣ was measured by use of an in-house RIA
with an antiserum (S781) raised against free hCG␣ and
affinity-adsorbed on a column of intact hCG conjugated to
CNBr-activated Sepharose. The tracer was 125I-radiolabeled purified free hCG␣ (NIH preparation CR123 ␣-subunit). The calibration curve ranged from 62.5 to 6250
pmol/L (0.9 –90.6 g/L) (19 ).
Free hCG was measured with an in-house IRMA
using antiserum (S752) raised in sheep against free hCG
and conjugated to 1,1⬘-carbonyldiimidazole-activated cellulose. Captured hCG was detected by an 125I-radiolabeled monoclonal antibody, 1/07, which recognizes
epitopes on the hCG C-terminal peptide. The calibration
curve used hCG (NIH preparation C123) and ranged
from 22 to 1110 pmol/L (0.5–24.6 g/L) (19 ).
The in-house hCGcf RIA used a polyclonal antibody
to hCGcf (S504) in a late-addition competition assay
using 125I-radiolabeled purified hCGcf. The calibration
curve ranged from 9 to 500 pmol/L (90 –5000 g/L) (20 ).
Data were summarized using box-and-whisker plots
and ROC curves. The Kruskal–Wallis test was used for
comparisons of concentrations. A P value of 0.05 was
2047
Clinical Chemistry 49, No. 12, 2003
regarded as the upper limit of significance. Statistical
analysis was performed with the Stats-Direct program,
and graphs were plotted using Origin 6.0
Results
The mean gestational ages (weeks) in women with ectopic
pregnancy, viable pregnancy, and spontaneous miscarriage were 7.0, 8.0, and 8.3 weeks, respectively; thus, no
significant difference in gestational age was found among
the groups. Table 1 shows the descriptive statistics for the
concentrations of hCG and its isoforms in serum and
urine sample for each patient group. The corresponding
figure can be seen in the Data Supplement that accompanies the online version of this article at http://www.
clinchem.org/content/vol49/issue12/. The Kruskal–Wallis test for nonparametric comparison of the patient
groups showed that, generally, concentrations of hCG
and its subunits were lower in ectopic pregnancy than in
normal pregnancies to a statistically significant margin
(see Table 2 in the online Data Supplement). The greatest
difference was found in serum free hCG in ectopic
pregnancy vs viable pregnancy (P ⬍0.0001). By way of
contrast, the concentrations of serum and urinary free
hCG␣ were not statistically different for ectopic pregnancies and viable pregnancies (P ⫽ 0.07 and 0.48, respectively; Table 1; also see Table 2 in the online Data
Supplement). However, women with early pregnancy
losses by spontaneous miscarriage also had dramatically
lower concentrations of the hCG isoforms and consequently could not be distinguished from those with
ectopic pregnancy (P ⫽ 0.45 for serum total hCG and P ⫽
0.70 for serum free hCG; see Table 2 in the online Data
Supplement).
The ROC curves for urinary and serum intact hCG, free
hCG, total hCG, free hCG␣, and hCGcf are shown in
Fig. 1. These curves show the likelihood of an ectopic
pregnancy determined by the different analytes below
optimal cutoff limits compared with a viable pregnancy
(Table 3 in the online Data Supplement). Serum free hCG
and serum total hCG are the best hCG-related analytes
and biological fluid for this purpose. Estimated areas
under the curves (95% confidence intervals) were 0.91
(0.72–1.0) and 0.86 (0.65–1.0) for serum free hCG and
total hCG, respectively. The reason for this was not
simply that they showed the greatest mean decrease in
concentrations: When we compared the decrease in group
median values, the serum analytes intact hCG and free
hCG had the lowest values, 12% and 13%, respectively,
of the values in women with viable pregnancies. However, it was the variability in the concentrations that
determined which analyte had superior diagnostic characteristic (Table 1; see also Tables 2 and 3 in the online
Data Supplement). The range of values for each analyte
was large, and most data sets did not follow a gaussian
distribution. Serum free hCG and total hCG had the
Table 1. Descriptive statistics for concentrations of hCG and related fragment molecules in serum and urine samples from
women with viable pregnancies, ectopic pregnancies, or spontaneous miscarriage.
Statistics
LMP,a weeks
Urinary isoforms,
pmol/mmol creatinine
Total hCG
Intact hCG
Free hCG
Free hCG␣
hCGcf
Serum isoforms, pmol/L
Total hCG
Intact hCG
Free hCG
Free hCG␣
a
Viable pregnancies
Mean (SD)
Range
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
Mean (SD)
Median (interquartile
LMP, last menstrual period.
b
Interquartile range is 75th–25th centile.
8.2 (2.3)
6.8–10
range)b
range)
range)
range)
range)
range)
range)
range)
range)
8109 (10 824)
2874 (10 982–1599)
22 814 (35 875)
8613 (26 246–3408)
4007 (5958)
1481 (4233–563)
33 407 (144 020)
3519 (6632–947)
17 477 (27 201)
12 753 (20 336–1034)
8909 (23 299)
1548 (5891–688)
25 737 (200 813)
203 993 (480 245–39 231)
936 (939)
645 (989–357)
3171 (3714)
2664 (4329–1132)
Ectopic pregnancies
7 (1.2)
6.7–8.0
1843 (3501)
716 (1591–178)
4499 (8041)
1649 (4395–660)
1698 (2020)
820 (3300–83)
26 054 (70 693)
1317 (5914–404)
4017 (4545)
2443 (7847–349)
360 (271)
284 (624–153)
59 664 (85 175)
24 719 (88 050–6134)
115 (90)
82 (195–35)
1231 (1177)
793 (2314–167)
Spontaneous miscarriage
9 (2.8)
6.8–11
1320 (2091)
657 (1743–227)
4675 (8447)
890 (5651–188)
895 (1769)
299 (846–48)
3457 (5657)
850 (3110–251)
2440 (3255)
920 (3137–398)
7988 (26 813)
430 (1634–123)
38 900 (67 473)
16 977 (39 953–1153)
188 (474)
69 (129–34)
1027 (1105)
343 (1948–167)
2048
Borrelli et al.: hCG Isoforms and Ectopic Pregnancy
Fig. 1. ROC curves for the hCG isoforms in
urine (A) and serum (B).
Solid light gray line, total hCG; solid black line,
intact hCG, dashed black line, hCG; dotted light
gray line, hCG␣; dot-dashed medium gray line,
hCGcf.
tightest distribution of values around the median (and
mean) values and consequently gave the best results.
Discussion
The validity of the current diagnostic algorithm for ectopic pregnancy, i.e., ultrasound and serial measurements
of hCG, has already been demonstrated by a decrease in
the risk of tubal rupture and morbidity of women whose
diagnosis was achieved by this protocol (21 ). Nevertheless, there are factors that reduce the sensitivity of ultrasound, such as lack of experience by the ultrasonographer, obesity of the patient, uterine myoma, or more
commonly, “inconclusive ultrasound”, which then must
be repeated several times (22 ). The subjective nature of
scanning contributes to this uncertainty. When considering these drawbacks to a diagnostic algorithm, it is clear
that a single blood test that would help to separate
patients with high risk for an ectopic pregnancy from
low-risk patients could possibly advance clinical practice
and be less expensive. This study was designed to examine whether the answer lay within the isoforms of hCG
produced by an ectopic pregnancy.
We found that concentrations of hCG and its subunits
were significantly lower in women with an ectopic pregnancy than in those with a normal pregnancy. There was,
however, considerable variability in individual patient
results, and serum results were far better discriminators
than urinary concentrations, even when corrected for
creatinine concentration. Nevertheless, certain hCG-related isoforms were clearly better than others. Serum free
hCG was ⬍500 pmol/L (⬍10 ng/mL) in all of the study
women presenting with ectopic pregnancies but ⬎500
pmol/L in the vast majority of those women with viable
pregnancies. This is consistent with the findings of Holman et al. (23 ), who found that all women with ectopic
pregnancies in their study had hCG concentrations ⬍25
g/L (1250 pmol/L). ROC analysis demonstrated up to
100% sensitivity for ectopic pregnancy with a 21% falsepositive rate (viable pregnancy) for serum free hCG at a
cutoff of 281 pmol/L (6.5 ng/mL). Total hCG performed
well when measured in both serum and urine samples,
but again serum measurements were best, giving 100%
sensitivity with a 32% false-positive rate (viable pregnancy) at a cutoff of 1053 pmol/L (375 IU/L). The data
would suggest that free hCG might have marginally
better performance characteristics with respect to the
diagnosis of ectopic pregnancy than total hCG. However,
the confidence intervals of the areas under the ROC
curves for these two measurements overlapped to such
an extent that the difference in performance was not
statistically significant. A much larger study is required to
determine whether there is a significant difference in the
diagnostic accuracy of free hCG and total hCG for
ectopic pregnancy.
Measurement of intact hCG and free hCG␣ gave less
significant results (Table 3 in the online Data Supplement), a finding inconsistent with that of Barnea et al.
(24 ), who reported that in combination with total hCG,
free hCG␣ could be used as a marker of ectopic pregnancy. Our results are also in disagreement with those of
Reuter et al. (25 ), who found that free hCG␣ concentrations were unchanged in ectopic pregnancies and increased in spontaneous abortions. No urinary analyte
apart from hCGcf was found to be a good predictor of
ectopic pregnancy. Cole et al. (26 ) proposed that hCGcf
could predict up to 84% of all ectopic pregnancies with
only 1% false positives. In our study, although hCGcf
measurements reached 100% sensitivity at a very high
cutoff of 14 201 pmol/L, this was at the expense of a very
poor specificity of 48% (see Table 3 in the online Data
Supplement).
Although serum free hCG and total hCG concentrations resolved ectopic pregnancy from viable pregnancy
with a high degree of diagnostic efficiency, spontaneous
miscarriage could not be distinguished. Because ⬃50% of
women presenting with suspected ectopic pregnancy will
subsequently be diagnosed with spontaneous miscarriage, the clinical utility of a single hCG measurement is
limited. Other markers, such as serum progesterone measurements, have been used to distinguish viable from
Clinical Chemistry 49, No. 12, 2003
nonviable (spontaneous miscarriage) pregnancies but
alone are unable to resolve ectopic from viable intrauterine pregnancies (22, 27 ).
To date, no single-point cutoff value for hCG has been
proposed as a discriminator of ectopic pregnancy. This
pilot study suggests that quantitative measurement of free
hCG in serum may provide this single-point cutoff for
resolving women at high risk from those with low risk of
ectopic pregnancy. Serum total hCG gave diagnostic
discriminator characteristics almost identical to those of
serum free hCG. This may be because total hCG assays
measure both free hCG and intact hCG, with the free
hCG component being the discriminating factor. The
study group in this report is too small to show any
statistical difference between the ability of total hCG and
free hCG to discriminate high-risk from low-risk suspected ectopic pregnancies.
In conclusion, based on our results, a single-point measure of total hCG or fee hCG could potentially improve
clinical practice in the diagnosis of ectopic pregnancy by
defining low risk of ectopic from high risk of ectopic
pregnancy for those women presenting at emergency
departments with lower abdominal pain and/or vaginal
bleeding and with a positive pregnancy test. A much
larger study is underway in a clinical setting, using
hospital-based automated assay platforms to compare the
abilities of total hCG and free hCG to differentiate
ectopic from viable pregnancies. This, in combination
with ultrasound and other serum markers that distinguish ectopic pregnancy from spontaneous miscarriage,
may be a more useful test for ectopic pregnancy than
current diagnostic algorithms.
References
1. Ectopic pregnancy—United States, 1990 –1992. MMWR Morb
Mortal Wkly Rep 1995;44:46 – 8.
2. Luciano AA, Roy G, Solima E. Ectopic pregnancy: from surgical
emergency to medical management. Ann N Y Acad Sci 2001;943:
235–54.
3. Kadar N. Early recourse to laparoscopy in the management of
suspected ectopic pregnancy. J Reprod Med 1990;35:1153– 6.
4. Braunstein GD, Ash RH. Predictive value analysis of measurements of hCG, pregnancy specific -1-glycoprotein, placental
lactogen and cysteine aminopeptidase for the diagnosis of ectopic
pregnancy. Fertil Steril 1983;39:62–7.
5. Tornehave D, Chemnitz J, Westergaard JG, Teisner B, Poulsen HK,
Bolton AE, et al. Placental proteins in peripheral blood and tissues
in ectopic pregnancies. Gynecol Obstet Invest 1986;23:97–102.
6. Stabile I, Olajide F, Chard T, Grudzinskas JG. Maternal serum
␣-fetoprotein levels in ectopic pregnancy. Hum Reprod 1989;4:
835– 6.
7. Cacciatore B, Ylostalo P, Stenman UH, Widholm O. Suspected
ectopic pregnancy: ultrasound findings and hCG levels assessed
by an immunofluorometric assay. Br J Obstet Gynaecol 1988;95:
497–502.
8. Ankum WM. Diagnosing suspected ectopic pregnancy. BMJ 2000;
321:1235– 6.
9. Frits MA, Guo SM. Doubling time of hCG in early normal pregnancy:
View publication stats
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
2049
relationship to hCG concentration and gestational age. Fertil Steril
1987;47:584 –9.
Shepherd RW, Patton PE, Novy MJ, Burry KA. Serial data hCG
measurements in the early detection of ectopic pregnancy. Obstet
Gynecol 1990;75:417–20.
Kadar N, De Vore G, Romero R. The discriminatory hCG zone. Its
use in the sonographic evaluation of ectopic pregnancy. Obstet
Gynecol 1981;58:156 – 61.
Aleem FA, De Fazio M, Gintautas J. Endovaginal sonography for
the early diagnosis of intrauterine and ectopic pregnancies. Hum
Reprod 1990;5:755– 8.
Di Marchi JM, Kosasa TS, Hale RW. What is the significance of
human chorionic gonadotropin value in ectopic pregnancy? Obstet
Gynecol 1989;74:851–5.
Bellisario R, Carlsen RB, Bahl OP. Human chorionic gonadotropin.
Linear amino acid sequence of the ␣-subunit. J Biol Chem
1973;248:6796 – 809.
Carlsen RB, Bahl OP, Swaminathan N. Human chorionic gonadotropin. Linear amino acid sequence of the -subunit. J Biol Chem
1973;248:6810 –27.
Iles RK, Chard T. Molecular insights into the structure and
function of human chorionic gonadotrophin. J Mol Endocrinol
1993;10:217–34.
Cole LA. Immunoassay of human chorionic gonadotropin, its free
subunits and metabolites. Clin Chem 1997;43:2233– 43.
Birken S, Berger P, Bidart JM, Weber M, Bristow A, Norman R, et
al. Preparation and characterization of new WHO reference reagents for human chorionic gonadotropin and metabolites. Clin
Chem 2003;49:144 –54.
Iles RK, Javid MK, Gunn LK, Chard T. Cross-reaction with luteinizing hormone -core is responsible for the age-dependent increase of immunoreactive -core fragment of human chorionic
gonadotropin in women with nonmalignant conditions. Clin Chem
1999;45:532– 8.
Lee CL, Iles RK, Shepherd JH, Hudson CN, Chard T. The purification and development of a radioimmunoassay for -core fragment
of human chorionic gonadotrophin in urine: application as a
marker of gynaecological cancer in premenopausal and postmenopausal women. J Endocrinol 1991;130:481–9.
Koh GH, Yeo GS. Diagnosis of ectopic pregnancy. Why we need a
protocol. Singapore Med J 1997;38:369 –74.
Gracia CR, Barnhart KT. Diagnosing ectopic pregnancy: decision
analysis comparing six strategies. Obstet Gynecol 2001;93:464 –
70.
Holman JF, Tyrey EL, Hammond CB. A contemporary approach to
suspected ectopic pregnancy with use of quantitative and qualitative assays for the -subunit of human chorionic gonadotropin
and sonography. Am J Obstet Gynecol 1984;150:151–7.
Barnea ER, Oelsner G, Benveniste R, Romero R, De Cherney AH.
Progesterone, estradiol and ␣-human chorionic gonadotropin secretion in women with ectopic pregnancy. J Clin Endocrinol Metab
1986;62:529 –31.
Reuter AM, Gaspard U, Deville JC, Vrindts-Gevaertv Y, Franchimont P. Evolution of serum levels of hCG and its subunits in
normal pregnancy and in threatened and spontaneous abortion.
Acta Endocrinol 1977;(Suppl):212–32A.
Cole LA, Kardana A, Seifer DB, Bohler HC Jr. Urine hCG -subunit
core fragment, a sensitive test for ectopic pregnancy. J Clin
Endocrinol Metab 1994;78:497–9.
Mol BWJ, Lijimer JG, Ankum WM, van der Veen F, Bossuyt PMM.
The accuracy of single serum progesterone measurement in the
diagnosis of ectopic pregnancy: a meta-analysis. Hum Reprod
1998;13:3220 –7.