FIELD OF THE INVENTION The present invention is in the field of prenatal diagnosis. It concerns non-invasive methods to screen prenatally for fetal Down syndrome and other aneuploidies by determining the levels of urinary gonadotropin peptide (UGP) [also known as β-core, β-core fragment, β-core-hCG, or urinary gonadotropin fragment (UGF) ] in maternal urine samples, alone or in conjunction with other markers.

BACKGROUND OF THE INVENTION Trisomy 21, commonly known as Down syndrome, is characterized by an extra copy of chromosome 21 and is one of the most common serious congenital abnormalities. People afflicted with Down syndrome have severe mental retardation, reduced life expectancies, and abnormal immune responses that predispose them to serious infections as well as thyroid autoimmunity. Further, 40% of Down syndrome patients have congenital heart disease and a 10 to 20-fold increased risk of developing leukemia over the general population. All Down syndrome patients older than 40 develop neuropathological changes characteristic of Alzheimer's disease.

Prenatal tests to detect fetal aneuploidies, such as Down syndrome (trisomy 21) , Turner syndrome (monosomy X) , Klinefelter syndrome (47, XXY), triple-X (47, XXX) among other aneuploidies, by amniocentesis or chorionic villus sampling (CVS) have been available since the late 1960s. Amniocentesis is the most common invasive prenatal diagnostic procedure. In amniocentesis, amniotic fluid is sampled by inserting a hollow needle through the mother's anterior abdominal and uterine walls into the amniotic cavity by piercing the chorion and amnion. It is usually performed in the second trimester of pregnancy. CVS is performed primarily during the first trimester, and involves collecting cells from the chorion which develops into the placenta.

Another invasive prenatal diagnostic technique is cordocentesis or percutaneous umbilical cord blood sampling, commonly known as fetal blood sampling. Fetal blood sampling involves obtaining fetal blood cells from vessels of the umbilical cord, and is often performed about the 20th gestational week.

Amniocentesis is used selectively because it presents a risk of about 1% of inducing spontaneous abortion. CVS and fetal blood sampling carry a similar or higher risk of inducing abortion, and there is also concern that those procedures may lead to fetal limb malformations in some cases. Thus, amniocentesis, CVS and fetal blood sampling are procedures that are only employed if a pregnancy is considered at high risk for a serious congenital anomaly. Thus, some means is required to select those pregnancies that are at a significant risk of Down syndrome or other fetal aneuploidies to justify the risks of such invasive prenatal diagnostic procedures, as amniocentesis, CVS and fetal blood sampling. Prior to 1983, the principal method for selecting pregnancies that had an increased risk for genetic defects, such as Down syndrome, was based on maternal age, that is, the older the age of the mother, the higher the risk that the pregnancy would be affected by aneuploidy. In 1974, biochemical screening for neural tube defects by measuring alpha-fetoprotein (AFP) began. In 1984, the use of the AFP screen was additionally adopted for the detection of Down syndrome. Since the early 1990s, a multiple marker blood test has been used to screen for this disorder. A common version of this test is the three marker triple test. The triple screen measures AFP, human chorionic gonadotropin (hCG) and unconjugated estriol (uE ₃ ) in the serum of pregnant women. The triple screen provides a means to screen the population of pregnant women to determine which pregnancies are at risk for Down syndrome and other serious genetic defects. The risk is calculated based on the results of the screen, along with other cofactors, such as, maternal age, to determine if the risk is high enough to warrant an invasive diagnostic procedure, such as, amniocentesis, CVS or fetal

blood sampling. Such prenatal screens, as the triple screen, can be used either to reduce the need for amniocentesis or to increase Down syndrome detection for the same amount of amniocentesis. "The efficiency of the Triple test is projected to be one case of fetal Down syndrome detected for every 50 amniocenteses performed." [Canick and Knight, "Multiple-marker Screening for Fetal Down Syndrome," Contemporary OB/GYN. pp. 3-12 (April 1992).]

Although pregnant women who are 35 years or older are the standard high risk group for fetal Down syndrome, screening also needs to be applied to young women because although they are at lower risk, most affected pregnancies are in young women. Approximately 80% of babies born with Down syndrome are born to mothers under 35. ["Down Syndrome Screening Suggested for Pregnant Women under 35," ACOG Newsletter. 38(8): 141 (Aug. 1994).]

The triple screen combines the analysis of three markers from serum to reduce false positive results (which result in the performance of unnecessary invasive procedures) and false negatives (in which serious genetic defects, such as, trisomy 21, go undetected) . In women under 35, the double screen (AFP and hCG) can pick up about half of Down syndrome cases and a large proportion of other chromosome defects during the second trimester. The triple screen (AFP, hCG and uE ₃ ) increases the detection rate by 5-10% of Down syndrome and a further increase in the detection of all other serious chromosome defects, thus decreasing the number of false- positives. Such rates mean that the double and triple screens still fail to detect a significant number of Down syndrome affected pregnancies.

Although the triple screen has a suggested screening period of 15 to 20 weeks gestation, such screening has been recommended between weeks 16-18 to maximize the window for spinal bifida detection. [Canick and Knight, supra (April 1992).] A 1992 survey of prenatal maternal serum screening for AFP alone or for multiple analyses reported that very few such screenings occurred in the thirteenth or earlier week of gestation. [Palomaki et al., "Maternal Serum Screening for

Fetal Down Syndrome in the United States: A 1992 Survey," Aπ . J. Obstet. Gvnecol.. 169(6): 1558-1562 (1992).] The triple screen thus suffers from the additional problem that once a risk of Down syndrome is predicted, and amniocentesis or another invasive prenatal definitive diagnostic procedure is performed to diagnose Down syndrome, it is at an advanced date of gestation, when termination of a pregnancy can be more physically and emotionally trying for the mother, and when certain less traumatic abortion procedures, such as, vacuum curettage, may not be available.

The limitations of the triple screen and the adverse consequences of unnecessary, potentially harmful and expensive invasive prenatal diagnostic procedures, such as, amniocentesis, have led to a search for more discriminatory markers for prenatal Down syndrome screening. Of the maternal serum markers in routine use, human chorionic gonadotropin (hCG) is the most discriminatory. [Cuckle et al., Prenatal Diagnosis. 14: 953-958 (1994).]

Human chorionic gonadotropin (hCG) is a glycopeptide hormone produced by the syncytiotrophoblasts of the fetal placenta, and has a molecular weight of about 38 kilodaltons (kd) . It can be detected by immunoassay in the maternal urine within days after fertilization and thus provides the basis of the most commonly used pregnancy tests. The intact hCG molecule is a dimer comprising a specific β subunit non- covalently bound to an a subunit, which is common to other giycoproteins.

Maternal serum levels of both intact hCG and the free β-subunit are elevated on average in Down syndrome but the extent of elevation is greater for free β-hCG [Spencer, K., Clin. Chem.. 37: 809-814 (1991); Spencer et al., Ann. Clin. Biochem.. 29: 506-518 (1992); Wald et al., Br. J. Obstet. Gvnaecol.. 100: 550-557 (1993)]. HCG is detected in the serum and urine of pregnant women, as are the free α- and β-subunits of hCG, and degradation products of hCG and of free β-subunit hCG.

The terminal degradation product of the β-subunit of hCG is called urinary gonadotropin peptide (UGP) , or

alternatively β-core-hCG, β-core fragment, β-core, or urinary gonadotropin fragment (UGF) . UGP is excreted into urine [Nislua et al., J. Steroid. Biochem.. 33: 733-737 (1989); Cole et al., J. Clin. Endocrinol. & Metab.. 76: 704-710 (1993)].

Urinary gonadotropin peptide (UGP) has an amino acid sequence related to the β-subunit of hCG. UGP is comprised of β-subunit residues 6 through 40 attached by disulfide linkages to residues 55 through 92. UGP is glycosylated but lacks the sialic acid and O-linked carbohydrate residues present on hCG β subunit.

UGP has been found in the urine of pregnant women carrying normal fetuses, and also in the urine of patients with gestational trophoblastic and non-trophoblastic malignancies [Cole et al., "Urinary Human Chorionic

Gonadotropin Free B-subunit and B-core Fragment: A New Marker of Gynecological Cancers," Cancer Res.. 48: 1356-1360 (1988); Cole et al., "Urinary Gonadotropin Fragments (UGF) in Cancers of the Female Reproductive System," Gynecol. Oncol.. 31: 82-90 (1988); O'Connor et al., "Development of Highly Sensitive

Immunoassays to Measure Human Chorionic Gonadotropin, Its β- subunit, and β-core Fragment in the Urine: Application to Malignancies," Cancer Res.. 48: 1361-1366 (1988); and Akar et al, "A Radioimmunoassay for the Core Fragment of the Human Chorionic Gonadotropin β-subunit," J. Clin. Endocrinol. and Metab.. 66: 538-545 (1988).] UGP has also been found to be associated with certain ovarian cancers. [Cole and Nam, "Urinary Gonadotropin Fragment (UGF) Measurements in the Diagnosis and Management of Ovarian Cancer," Yale J. Bio, and Med.. 62: 367-378 (1989).]

Cuckle et al., Prenatal Diagnosis. 14: 953-958 (1994) showed that UGP levels are elevated on average in the second trimester of singleton pregnancies affected by fetal Down syndrome and reduced in the first trimester of singleton pregnancies affected with other serious, but less common aneuploidies such as, Edwards syndrome and triploidy, and of a twin pregnancy discordant for Down syndrome. The observed median level in Down syndrome (6.11 MOM: 95% confidence

interval 3.7 to 10.0) was greater than the corresponding median level for intact hCG in maternal serum (2.0 MOM; 1.9- 2.1) and free β-hCG (2.3 MOM; 2.1-2.5).

There are important advantages to using urinalysis for prenatal screening for Down syndrome. Urine tests are less expensive than serum testing, avoid the safety issues and handling risks associated with the collection and storage of blood samples, as well as the invasiveness and discomfort of phlebotomy. Urine samples can be easily collected and shipped, if necessary, where women have limited access to medical testing facilities because of geography or socio- economic status. UGP is stable to changes in temperature, pH, and storage time at -20 and 40°C.

However, the UGP assay results described in Cuckle et al., supra showed elevated UGP levels in maternal urine samples taken between the 19th week and the 22nd week plus 4 days of gestation from Down syndrome affected pregnancies. As indicated above, there are disadvantages to second trimester testing, in that delays in confirming a fetal Down syndrome diagnosis result in more traumatic abortion procedures being necessitated. Also, the emotional attachment and expectations of the pregnant woman and her family for a healthy baby, grow during the pregnancy, making the abortion decision more difficult later in the gestational term. The instant invention provides the benefits of urinalysis and avoids the problems of second trimester prenatal screening by providing methods to screen first trimester urine samples for fetal Down syndrome and other aneuploidies. According to the methods of this invention, a UGP level in a first trimester maternal urine sample when elevated above normal levels indicates a significant risk of the pregnancy under analysis being affected by Down syndrome or another fetal aneuploidy, such as, Turner syndrome, Klinefelter syndrome or triple-X. Preferred prenatal screening methods of the instant invention are highly specific for UGP and minimally cross-reactive with intact hCG, with β- subunit hCG and with α-subunit hCG.

SUMMARY OF THE INVENTION The instant invention provides methods for prenatally determining whether there is a significant risk of a pregnancy being affected by Down syndrome or other fetal aneuploidies by testing first trimester maternal urine samples for elevations of urinary gonadotropin peptide (UGP) levels above normal. Antibody and non-antibody methods are disclosed to detect and quantitate UGP levels in maternal urine samples. One representative embodiment of the invention employs immunoassays that are specific for UGP and have molar cross-reactivities of less than about 10% with intact hCG, with β-subunit hCG, and with α-subunit hCG. More preferably such immunoassay methods of this invention have a molar cross¬ reactivity of less than about 5%, more preferably less than about 3%, and still more preferably less than about 1%, with intact hCG, with β-subunit hCG, and with α-subunit hCG.

The UGP level in a first trimester maternal urine sample is related to the median for unaffected pregnancies at approximately the same gestational age, wherein a UGP level above that median indicates a risk of Down syndrome or another fetal aneuploidy such as, Turner syndrome, Klinefelter syndrome or triple-X, affecting the pregnancy under analysis, the degree of elevation above that median being considered. Results from the prenatal screening methods of this invention are generally expressed as multiples of the gestation-specific median value (MOM) for unaffected pregnancies of the same gestation. Exemplary positive results from the screening methods according to this invention are those wherein the UGP level is from about 1.1 MOM to higher multiples, from about 1.5 MOM to higher multiples, from about 2 MOM and higher multiples.

The UGP screening result from the methods of this invention is used to assess the fetal Down syndrome or other fetal aneuploidies either alone or in conjunction with results from other screening tests with other serum and/or urinary markers, and/or other factors, such aε, maternal age, maternal health, maternal weight and ultrasound parameters among other factors. For example, maternal age and UGP levels are

independent predictors of Down syndrome risk, as is true for each of the commonly used serum markers. Therefore, after performing the prenatal screening methods of this invention, the risk of a Down syndrome affected pregnancy can be calculated by multiplying the age-related risk by a likelihood ratio derived from the UGP level found in the maternal urine sample in relation to control samples.

Other urinary markers which could be preferred for assessing the risk of a Down syndrome affected pregnancy in conjunction with UGP levels, include pregnancy-associated plasma protein A (PAPP-A) , dimeric inhibin, total estrogen (tE) , unconjugated estriol (uE ₃ ) , total estriol (tE ₃ ) , AFP and proform of eosinophilic major basic protein (proMBP) , among other urinary marker possibilities. In general, a positive result from the screening methods of this invention is an indicator that a more invasive prenatal diagnostic procedure, such as, amniocentesis, CVS or fetal blood sampling, should be performed to determine definitively whether the pregnancy is affected with Down syndrome or another fetal aneuploidy.

Gestation-specific medians for UGP can be calculated by weighted non-linear regression from the values for control urine samples. To account for variations in the concentrations of urine samples, UGP levels can be expressed in terms of creatinine. Gestational ages of cases and controls can be determined by ultrasound parameters and by last menstrual period dating.

The control samples are preferably taken from a population of pregnant women that are matched as well as practicable to the population from which the pregnant woman who provided the test sample comes. For example, population parameters could include race, ethnicity, and geographical location, among other parameters.

The immunoassay methods of this invention employ UGP standards. Blithe et al., U.S. Patent No. 5,445,968 (issued August 29, 1995) discloses methods of purifying UGP.

The prenatal screening methods of this invention using immunoassays can be in any standard immunoassay format,

for example, a competitive radioimmunoassay, a sandwich EIA or sandwich radioimmunoassay (RIA), among other known formats. A sandwich assay is a preferred format of this invention, and a sandwich EIA is a further preferred embodiment. The prenatal screening methods of this invention can be automated. A preferred automated immunoassay system is Ciba Corning Diagnostic Corp.'s (CCD's) ACS-180™ Automated Chemiluminescence System [CCD; Medfield, MA (USA) ] .

The prenatal screening methods of this invention employ antibodies, that are defined herein to include whole antibodies or biologically active fragments of antibodies. The antibodies used in the immunoassay methods can be monoclonal and/or polyclonal, preferably monoclonal and/or affinity-purified polyclonal antibodies. The specificity of the immunoassay methods is provided by antibodies which specifically bind to UGP; generally such antibodies are monoclonal antibodies.

Tracer antibodies that can be used in the immuno¬ assay methods of this invention can be directly or indirectly linked to a detectable marker. The signal from said marker can indicate the level of UGP in the sample tested. The signal's intensity may be directly proportional to the level of UGP in the sample.

Exemplary detectable markers can be selected from the group consisting of radionuclides, fluorescers, bioluminescers, chemiluminescers, dyes, enzymes, coenzymes, enzyme substrates, enzyme cofactors, enzyme inhibitors, enzyme subunits, metal ions, and free radicals.

Antibodies used in the immunoassay methods may be linked to a solid phase, for example, the wall of a container or the surface of magnetic or paramagnetic particles, among other solid phases.

DETAILED DESCRIPTION The following abbreviations are used herein:

Abbreviations

AFP - alpha-fetoprotein

α-hCG free α-subunit of human chorionic gonadotropin

BL bioluminescent

CCD Ciba Corning Diagnostics Corp.

CL chemiluminescent creat. creatinine

CVS chorionic villus sampling

EIA enzyme immunoassay fmol femtomole gest. gestational hCG human chorionic gonadotropin

HRP horseradish peroxidase

L liter

LNMP last normal menstrual period

MBP major basic protein g milligram ml milliliter mmol millimole

MOM multiples of normal gestation-specific median ng nanogram

PAPP-A pregnancy-associated plasma protein A pmol picomole

PMP paramagnetic particle proMBP proform of eosinophilic major basic protein

RIA radioimmunoassay

SD standard deviation tE total estrogen tE ₃ total estriol

TMB tetramethyl benzidine

UE, unconjugated estriol UGF urinary gonadotropin fragment UGP urinary gonadotropin peptide

Definitions

Alternative terms used in the art for urinary gonadotropin peptide (UGP) are β-core fragment, β-core-hCG, and urinary gonadotropin fragment (UGF) .

The first trimester is herein defined as 14 completed weeks (14 weeks, 6 days) from the onset of a pregnant woman's last normal menstrual period (LNMP) .

Intact hCG is a term that defines hCG in its dimeric form when its α and β subunits are non-covalently bound together.

Total hCG is a term that includes intact hCG and either its free α subunit or its free β subunit.

"Aneuploidy" is defined herein to refer to any deviation from the human diploid number of 46 chromosomes.

The term "antibodies" is defined herein to include not only whole antibodies but also biologically active fragments of antibodies, preferably fragments containing the antigen binding regions.

Representative Embodiments

Herein are disclosed methods for prenatally assessing risks of a pregnancy being affected by Down syndrome or other fetal aneuploidies by testing maternal urine samples for elevations above normal of urinary gonadotropin peptide (UGP) . The methods employ antibody and non-antibody procedures for detecting and quantitating UGP in such samples. Elevated UGP levels above normal in a maternal urine sample indicate a significant risk of Down syndrome or other aneuploidies, such as, Turner syndrome, triple-X or Klinefelter syndrome among other aneuploidies.

The incidence of Turner syndrome (monosomy X) in new-born females is about one in 5,000 live female births [Thompson et al., Thompson and Thompson Genetics in Medicine. (5th ed.; WB Saunders; Philadelphia, PA, USA; 1991)]. The incidence of triple-X (47, XXX) is one in 960 live births, and the incidence of Klinefelter syndrome (47, XXY) is one in 1080 live births [Hook and Hamerton, "The frequency of chromosome abnormalities detected in consecutive newborn studies . . . " IN: Population Cytogenetics: Studies in Humans [Hook and Porter (eds.); Academic Press; NY, USA (1977)].

One preferred embodiment of the prenatal screening methods of this invention is expressed as follows.

A method for prenatally determining whether there is a significant risk of a pregnancy being affected by fetal aneuploidy during the pregnancy's first trimester comprising:

(a) taking a maternal urine sample during the first trimester of said pregnancy;

(b) testing said maternal urine sample to determine the level of UGP in said sample;

(c) determining whether the level of UGP in said sample is elevated above a level of UGP that is normal in urine samples from women whose pregnancies are unaffected by aneuploidy, and whose pregnancies are at about the same gestational age as the pregnancy under analysis; and

(d) determining whether there is a significant risk of fetal aneuploidy based upon the level of UGP in said sample, wherein a level elevated above said normal level indicates that there is a significant risk of the pregnancy being affected by fetal aneuploidy.

Many methods can be used to detect and quantitate UGP in maternal urine samples. Antibody and non-antibody methods can be used. A variety of immunoassay formats can be used as set forth below. Non-antibody methods include, for example, chromatographic procedures that separate UGP from other components of urine, such as, high pressure liquid ■chromatography (HPLC) and mass spectrometry, alone or in combination; fluorometric detection means; nuclear magnetic resonance (NMR) ; and the use of non-antibody receptors and other binding proteins that may exist cellularly or in serum.

Prefereably an immunoassy is used to detect and quantitate UGP in maternal urine samples in the methods of this invention. Such immunoassays can be in any standard immunoassay format, such as, sandwich assays, competition assays, bridge immunoassays, among other formats well known to those of skill in the art. [See, for example, U.S. Patent Nos. 5,296,347; 4,233,402; 4,034,074; and 4,098,876.] For example, a competitive radioimmunoassay (RIA) , a sandwich EIA or sandwich RIA may be preferred immunoassay formats. A sandwich assay is a preferred format of this invention, and a sandwich RIA or EIA is a further preferred format.

A particularly preferred immunoassay to detect and quantitate UGP is the radioimmunoassay described herein in Example 4 [a modification of the method described in Lee et al., J. Endocrinol.. 130: 481-489 (1991)]. A further particularly preferred immunoassay for use in the methods of this invention is the Triton ^R UGP EIA Kit from Ciba Corning Diagnostics Corp. (Alameda, CA; USA) , which was used in Examples 1-3, infra.

In one representative embodiment of the instant invention, an immunoassay that is highly specific for UGP is used, which has a molar cross-reactivity of less than about 10% with each of the following: intact hCG, β-subunit hCG, and with α-subunit hCG. Preferably, such an immunoassay specific for UGP preferably has molar cross-reactivities of less than 5% with intact hCG, with β-subunit hCG, and with α- subunit hCG; more preferably those cross-reactivities are less than 3%; and still more preferably those cross-reactivities are less than 1%.

The Triton" UGP EIA Kit from Ciba Corning Diagnostics Corp. (CCD; Alameda, CA; USA) is such a preferred immunoassay that is highly specific for UGP. That commercially available immunoassy exhibits the following molar cross-reactivities: hCG, 0.11 per cent; free beta subunit of hCG, 0.043 per cent, and free alpha subunit of hCG, 0.009 per cent. Other representative immunoassays highly specific for UGP can comprise, for example, antibodies described in Blithe et al., U.S. Patent No. 5,445,868 (issued August 29, 1995).

In a preferred embodiment of the invention, monoclonal antibodies specific to UGP are coupled to paramagnetic particles (PMP) and incubated with the urine sample to be tested. The immobilized complex is washed to remove unbound materials. Anti-UGP antibodies conjugated with a chemiluminscent tag, preferably acridinium ester (AE) , are then incubated with the immunocomplex immobilized on the PMP. A step to separate bound tracer from the unbound tracer is then performed. The bound fraction is then incubated with the light reagent and the amount of light photons generated iε determined in a luminometer, such as, for example, the Ciba

Corning ACS-180™ Automated Chemiluminescent System [Ciba Corning Diagnostics Corp. (CCD) E. Walpole, MA (USA)], which is further described infra. The values obtained are directly proportional to the concentration of UGP in the urine sample. The embodiments outlined above are representative of a wide number of assay methods that can be used in accordance with this invention. There are many variations and modifications of such outlined embodiments within the skill of one in the art. Variations of the representative embodiments of the methods of this invention within conventional knowledge of those of skill in the art are considered to be within the scope of the instant invention. Preferred variations and more detailed embodiments are identified in the following sections. Reference is made hereby to standard textbooks of immunology that contain methods for carrying out various immunoassay formats that can be adapted from those specifically represented herein. See, for example, Molecular Biology and Biotechnology: A Comprehensive Desk Reference (Ed. R. A. Meyers) [VCH Publishers, Inc., New York, N.Y. (1995)]; Moore and Persaud, The Developing Human: Clinically Oriented Embryology. 5th Edition [W. B. Saunders Company; Philadelphia/London/Toronto/Montreal/Sydney/Tokyo (1993) ] ; Darnell et al. , Molecular Cell Biology. W. H. Freeman and Company (N.Yi 1990); Colowick et al., Methods in Enzymology. Volume 152 [Academic Press, Inc. (London) Ltd. (1987)]; and

Goding J. W. , Monoclonal Antibodies: Principles and Practice: Production and Application of Monoclonal Antibodies in Cell Biology. Biochemistry, and Immunology [Academic Press Inc. (London) Ltd.; 1983.]

Timing of Sample Taking

Generally, maternal urine samples can be taken for testing, according to this invention, during the fifth week to the 14th completed week, that is, 14 weeks plus six days, of gestation. Preferably, maternal urine samples can be taken for testing during the sixth week to the 14th completed week of gestation; more preferably during the seventh week to the 14th completed week of gestation; still more preferably during

the eighth week to the 14th completed week of gestation; further preferably, during the ninth week to the 14th completed week of gestation; alternatively, during the 10th week to the 14th completed week of gestation; further alternatively during the ninth week to the 13th week of gestation; and still further also preferably from the ninth week to the twelfth completed week of gestation.

Other preferred gestational periods from the first trimester, during which maternal urine samples can be taken for testing according to this invention include the following: the gestational periods from the beginning of the seventh week to the end of the 14th week; from the beginning of the eighth week to the end of the 14th week; from the beginning of the ninth week to the end of the 14th week; from the beginning of the tenth week to the end of the 14th week; from the beginning of the fifth week to the end of the 13th week; from the beginning of the fifth week to the end of the 12th week; from the beginning of the sixth week to the end of the 13th week; from the beginning of the sixth week to the end of the 12th week; from the beginning of the seventh week to the end of the 13th week; from the beginning of the seventh week to the end of the 12th week; from the beginning of the eighth week to the end of the 13th week; from the beginning of the eighth week to the end of the 12th week; from the beginning of the eighth week to the end of the 11th week; from the beginning of the eighth week to the end of the 10th week; from the beginning of the ninth week to the end of the 11th week; from the beginning of the ninth week to the end of the 10th week; from the beginning of the tenth week to the end of the 13th week; and from the beginning of the tenth week to the end of the twelfth week.

Urine Concentration

The prenatal screening methods of this invention do not require uniformity in the volume of urine obtained from each pregnant woman of the control or case groups or the time of voiding. The UGP concentration can be corrected for

creatinine, for example, expressed in mmol/mmol or nmol/mmol creatinine.

The CCD UGP kit expresses UGP levels as pmol per mg urinary creatinine. UGP concentrations may be reported in International Units as mmol UGP/L and expressed as nmol/mmol creatinine as exemplified in Example 2, infra.

Creatinine content in urine can be measured by the Jaffe method using a Monarch 200 centrifugal analyser as described in Cuckle et al.. Prenatal Diagnosis. 14: 953-958 (1994) . Creatinine levels can also be measured on a Synchron CX-5 chemistry analyser (Beckman Instruments; Brea, CA, USA) .

Other Markers

"Population screening is the identification, among apparently healthy individuals, of those who are sufficiently at risk of a specific disorder to justify a subsequent diagnostic test or procedure. This implies the testing of all pregnancies in order to identify those few at a great enough risk to warrant an invasive diagnostic procedure, such as amniocentesis." [Canick et al., "Maternal Serum Screening for Aneuploidy and Open Fetal Defects," Prenatal Diagnosis. 20(3): 443 (Sept. 1993).] A positive result from a screen indicates a higher risk of fetal Down syndrome, and a negative result does not necessarily mean that a fetus is free of Down syndrome but that the fetus is at a lower risk of the syndrome.

Screening detection rates can be increased and false positives can be decreased by combining the resultε of a prenatal screening method with results of screening with other markers and assessing the resultε in conjunction with other factorε. When considering the cost of a second or third marker, it has to be weighed against the extra costs incurred without their use in additional amniocentesis tests for the affected pregnancies. [Cuckle, H. S., Clin. Chem.. 38(9) : 1687-1689 (1992).] As indicated above, the results from the prenatal screening methods of this invention can be used alone or in conjunction with results from other screening tests with other

serum and/or urinary markers, and/or other factors, such as, ultrasound parameters, maternal age, maternal health, maternal weight, among other factors to assess the risks of a pregnancy being affected by Down syndrome or other fetal aneuploidies. For example, maternal age and UGP level are independent predictors of Down syndrome risk, as is true for each of the commonly used serum markers.

Other markers which could be preferred urinary markers for assessing the risks of a Down syndrome affected pregnancy include PAPP-A, dimeric inhibin, tE, tE ₃ uE ₃ , AFP, 16α-OH-DHAS and proMBP, among other urinary marker possibilities. Particularly preferred potential urinary markers to be used in conjunction with UGP to assess fetal Down syndrome risks are PAPP-A, dimeric inhibin and tE. Example 4 illustrates testing urine samples for UGP, tE and α-hCG. As illustrated in Example 4, the levels of tE and α-hCG in maternal urine samples when reduced below respective normal levels for those markers indicate a significant risk of fetal aneuploidy, such as, Down syndrome or Turner syndrome.

PAPP-A may be found in first trimester urine samples, preferably taken during the ninth to the 13th or 14th completed week of gestation, at levels significantly reduced below normal in Down syndrome affected pregnancies. An assay for dimeric inhibin is described in Cuckle et al. , "Maternal Serum Inhibin Levels in Second-Trimester Down's Syndrome Pregnancies," Prenatal Diagnosis. 14: 387-390 (1994) . That asεay could be adapted for urine teεting. Dimeric inhibin may be useful in the first trimester. Total eεtrogen can be meaεured by a continuouε flow εyεtem baεed on the Kober reaction [Lever et al., "Improved estriol determination in a continuous flow syεtem," Biochem. Med.. 8: 188-198 (1973)]. Total estrogen could be measured as μmol/mmol creatinine. By the second trimester, maternal estrogen is largely derived from fetal hepatic and placental metabolism of precursors secreted by the fetal adrenal glands [Oakey, R.E., Vitamins and Hormones. 28: 1 (1970)]. Thus, measurement of

estrogen production reflects fetal and placental function. The unconjugated estrogens formed by the placenta are rapidly converted to conjugated forms by addition of glucuronic and sulphuric acid residues, before excretion. The inventors measured those urinary conjugated estrogens, mostly estriol conjugates. By contrast, in maternal serum screening for Down syndrome, unconjugated estriol is meaεured rather than the more abundant conjugated formε.

The method uεed in Example 4 herein assesses excretion of a combination of 16α-hydroxyestrogenε (e.g. estriol) and deoxyeεtrogens (e.g. estrone, estradiol) . A different method would be to target one particular estrogen (e.g. , estriol-16α-glucosiduronate) .

Improved discrimination may be achieved by measuring precursors directly in the urine. A preferred precursor is 16α-hydroxydehydroepiandrosterone sulphate (16α-OH-DHAS) . Through a placental procesε known aε aromatization, this compound is the precursor of more than 85% of estrogens in late pregnancy urine. I6α-0H-DHAS has a relatively low affinity for the series of placental enzymes which catalyze its conversion to estriol [Purohit and Oakey, J. Ster. Biochem.. 33: 439-448 (1989)]. Any decline in placental function in Down syndrome might well exaggerate the quantity which escapes aromatization. In the extreme condition of steroid sulphatase deficiency, for example, sufficient 16α-0H- DHAS avoidε aromatization and iε excreted in maternal urine to provide a εecure diagnostic test for the condition [Wil ot et al., Ann. Clin. Biochem.. 25: 155-161 (1988)].

Alεo, phyεical measurements, such as ultraεound markerε, can be uεed in combination with UGP levelε to assess prenatally the risk of fetal aneuploidy. Exemplary ultrasound markerε include nuchal tranεlucency [Nicolaideε et al., Br. Med. J.. 304: 867 (1992)], femur length, hu eruε, cephalic index, ventricular εize, and ultrasound detected defects in the heart, gut, and organs, among other physical markers.

Antibodies

As indicated above, the term "antibodies" is defined to include not only whole antibodies but also biologically active fragments of antibodies, preferably fragments containing the antigen binding regions. Such antibodies may be prepared by conventional methodology and/or by genetic engineering.

Antibody fragments may be genetically engineered, preferably from the variable regions of the light and/or heavy chains (VH and VL) , including the hypervariable regions, and still more preferably from both the VH and VL regions. For example, the term "antibodies" as used herein comprehends polyclonal and monoclonal antibodies and biologically active fragments thereof including among other possibilities "univalent" antibodies [Glennie et al., Nature. 295: 712 (1982)]; Fab proteins including Fab' and F(ab')2 fragments whether covalently or non-covalently aggregated; light or heavy chains alone, preferably variable heavy and light chain regions (VH and VL regions) , and more preferably including the hypervariable regions [otherwise known as the complementarity determining regions (CDRs) of εaid VH and VL regionε] ; Fc proteins; "hybrid" antibodies capable of binding more than one antigen; constant-variable region chimeras; "composite" immunoglobulins with heavy and light chains of different origins; "altered" antibodies with improved specificity and other characteristicε aε prepared by standard recombinant techniqueε and also by oligonucleotide-directed mutagenesis techniques [Dalbadie- McFarland et al., PNAS (USA), 79_: 6409 (1982) ] . It may be preferred for many immunoassayε of this invention that biologically active fragments rather than whole antibodies be used. Fab fragments are particularly preferred fragments in accordance with this invention to avoid non¬ specific binding. Antibodies for use in the instant invention can be genetically engineered. [See, for example, Morrison et al., Clin. Chem.. 34: 1668 (1988) ; Morrison and Oi, Adv. Immunol. , 44: 65 (1989); Rodwell, Nature. 342: 99 (1989); Pluckthun,

A. , Nature. 347: 497 (1990); Winter and Milstein, Nature. 349: 293 (1991); Pluckthun, A., Bio/Technology. 9: 545 (1991); Wetzel, R. , Protein Eng.. 4: 371 (1991); Geisow, M.J., Trends Biotechnol.. 10: 75 (1992); and Chiswell and McCaffery, Trends Biotechnol. 10: 85 (1992).] Further bispecific and other types of antibodies [for example, Lerner and Tramanto, Trends Biochem. Sci.. 12: 427 (1987) ; Shokat and Schultz, Annu. Rev. Immunol.. 8: 335 (1990); Schultz, P. G. , Science. 240: 426 (1988); Benkavic et al., Science. 250: 1135 (1990); and Lerner et al., Science. 252: 659 (1991); Noland and 0'Kennedy, Biochem. Biophys. Acta.. 1040: 1 (1990); and Bolhuis et al., Cell Biochem.. 47: 306 (1991)] can also be used according to this invention.

Standards Purified reference preparations can be obtained by eεtabliεhed procedures. [See, for example. Blithe et al., "Purification of β-core fragment from pregnancy urine and demonstration that itε carbohydrate moietieε differ from thoεe of native human chorionic gonadotropin-β," Endocrinol.. 122: 173-180 (1988).]

Automated Immunoaεεav Svεtem

The methodε of thiε invention can be readily adapted to automated immunochemiεtry analyzers. To facilitate automation of the methods of this invention and to reduce the turnaround time, anti-UGP antibodies may be coupled to magnetizable particles.

A preferred automated/immunoassay system is the Ciba Corning ACS:180™ Automated Chemiluminescence System [CCD; Medfield, MA (USA)]. The Ciba Corning ACS:180™ Automated Immunoasεay Syεtem iε deεcribed in Dudley, B.S., J. Clin. Immunoaεεay. 14(2) : 77 (Summer 1991) . The system uses chemilumineεcent labels as tracers and paramagnetic particles as solid-phase reagents. The ACS:180 system accommodates both competitive binding and sandwich-type asεays, wherein each of the εteps are automated. The ACS:180 uses micron-sized paramagnetic particles that maximize the available surface

area, and provide a meanε of rapid magnetic separation of bound tracer from unbound tracer without centrifugation. Reagents can be added simultaneouεly or sequentially. Other tags, such as an enzymatic tag, can be used in place of a chemiluminescent label, such as, acridinium ester.

Solid Phase

The solid phase used in the assays of this invention may be any surface commonly used in immunoassays. For example, the solid phase may be particulate; it may be the surface of beads, for example, glaεs or polystyrene beadε; or it may be the εolid wall surface of any of a variety of containerε, for example, centrifuge tubes, columns, microtiter plate wells, filters, membranes and tubing, among other containerε. When particleε are uεed aε the solid phase, they will preferably be of a size in the range of from about 0.4 to 200 microns, more usually from about 0.8 to 4.0 microns. Magnetic or magnetizable particles such as, paramagnetic particles (PMP) , are a preferred particulate solid phase, and microtiter plate wellε are a preferred solid wall surface. Magnetic or magnetizable particles may be particularly preferred when the steps of the methods of this invention are performed in an automated immunoassay system.

Preferred detection/quantitation systems of this invention may be luminescent, and a luminescent detection/quantitation system in conjunction with a signal amplification system could be used, if necessary. Exemplary luminescent labels, preferably chemiluminescent labels, are detailed below, as are signal amplification systems.

Signal Detection/Ouantitation Svstemε

The complexeε formed by the assays of this invention can be detected, or detected and quantitated by any known detection/quantitation systems used in immunoassays. As appropriate, the antibodies of this invention used as tracers may be labeled in any manner directly or indirectly, that

resultε in a signal that is visible or can be rendered visible.

Detectable marker substanceε include radionuclides, such as ³ H, ¹²⁵ I, and ¹³¹ I; fluorescers, such as, fluorescein isothiocyanate and other fluorochromes, phycobiliproteins, phycoerythin, rare earth chelates, Texas red, dansyl and rhodamine; colorimetric reagents (chromogens) ; electron-opaque materials, such as colloidal gold; bioluminescers; chemiluminescers; dyes; enzymes, such aε, horseradish peroxidase, alkaline phosphatase, glucose oxidase, glucose-6-phosphate dehydrogenase, acetylcholinesterase, α-, β-galactosidase, among others; coenzymeε; enzyme εubεtrateε; enzyme cofactorε; enzyme inhibitors; enzyme subunits; metal ions; free radicals; or any other immunologically active or inert subεtance which provideε a means of detecting or measuring the preεence or amount of immunocomplex formed. Exemplary of enzyme εubεtrate combinationε are horseradish peroxidase and tetramethyl benzidine (TMB) , and alkaline phosphatase and paranitrophenyl phosphate (pNPP) . Preferred detection, or detection and quantitation systems according to this invention produce luminescent εignals, bioluminescent (BL) or chemiluminescent (CL) . In chemiluminescent (CL) or bioluminescent (BL) assays, the intensity or the total light emission is measured and related to the concentration of the analyte. Light can be meaεured quantitatively using a luminometer (photomultiplier tube as the detector) or charge-coupled device, or qualitatively by means of photographic or X-ray film. The main advantages of using such asεayε is their simplicity and analytical sensitivity, enabling the detection and/or quantitation of very small amounts of analyte.

Exemplary luminescent labels are acridinium esters, acridinium sulfonyl carboxamides, luminol, umbelliferone, isoluminol derivativeε, photoproteinε, εuch as aequorin, and luciferaseε from fireflieε, marine bacteria, Vargulla and Renilla. Luminol can be uεed optionally with an enhancer molecule, preferably εelected from the group consisting of 4-iodophenol or 4-hydroxy- cinnamic acid. Acridinium esters

are one of the preferred types of CL labels according to this invention. A signal is generated by treatment with an oxidant under basic conditions.

Also preferred luminescent detection systems are those wherein the signal is produced by an enzymatic reaction upon a εubstrate. CL and BL detection schemes have been developed for assaying alkaline phosphatase (AP) , glucose oxidase, glucose 6-phosphate dehydrogenase, horseradish peroxidase (HRP) , and xanthine-oxidaεe labelε, among others. AP and HRP are two preferred enzyme labels which can be quantitated by a range of CL and BL reactions. For example, AP can be used with a substrate, such as an adamantyl 1,2-dioxetane aryl phosphate substrate (e.g. AMPPD or CSPD; [Kricka, L.J. , "Chemiluminescence and Bioluminescence, Analysis by," at p. 167, Molecular Biology and Biotechnology: A Comprehenεive Deεk Reference (ed. R.A. Meyerε) (VCH Publiεherε; N.Y., N.Y.; 1995)]; preferably a disodium salt of 4-methoxy-4-(3-phosphatephenyl)εpiro [1,2-dioxetane- 3,2'-adamantane] , with or without an enhancer molecule, preferably, l-(trioctylphoεphonium methyl)-4-

(tributylphoεphonium methyl) benzene diochloride. HRP is preferably used with substrates, such as, 2',3',6'- trifluorophenyl 3-methoxy-10-methylacridan-9-carboxylate.

CL and BL reactions can also be adapted for analysiε of not only enzymes, but other substrates, cofactors, inhibitors, metal ions and the like. For example, luminol, firefly luciferase, and marine bacterial luciferase reactions are indicator reactions for the production or consumption of peroxide, ATP, and NADPH, respectively. They can be coupled to other reactions involving oxidases, kinases, and dehydrogenases, and can be used to measure any component of the coupled reaction (enzyme, subεtrate, cofactor) .

The detectable marker may be directly or indirectly linked to an antibody used in an assay of this invention. Exemplary of an indirect linkage of the detectable label is the use of a binding pair between the antibody and the marker, or the use of well known εignal amplification signals, such

as, using a biotinylated antibody complexed to UGP and then adding strepavidin conjugated to HRP and then TMB.

Exemplary of binding pairs that can be used to link antibodies of assays of this invention to detectable markers are biotin/avidin, streptavidin, or anti-biotin; avidin/anti-avidin; thyroxine/thyroxine-binding globulin; antigen/antibody; antibody/anti-antibody; carbohydrate/ lectins; hapten/anti-hapten antibody; dyes and hydrophobic moleculeε/hydrophobic protein binding sites; enzyme inhibitor, coenzyme or cofactor/enzyme; polynucleic acid/homologous polynucleic acid sequence; fluorescein/anti-fluorescein; dinitrophenol/ anti-dinitrophenol; vitamin B12/intrinsic factor; cortisone, cortisol/cortisol binding protein; and ligandε for εpecific receptor protein/membrane aεεociated specific receptor proteins. Preferred binding pairs according to this invention are biotin/avidin or εtreptavidin, more preferably biotin/ treptavidin.

Variouε means for linking labels directly or indirectly to antibodies are known in the art. For example, labels may be bound either covalently or non-covalently. Exemplary antibody conjugation methodε are described in: Avarmeas et al., Scan. J. Immunol.. 8 (Suppl. 7): 7 (1978); Bayer et al., Meth. Enzymol.. 62: 308 (1979); Chandler et al., J. Immunol. Meth.. 53: 187 (1982); Ekeke and Abuknesha, J. Steroid Biochem.. 11: 1579 (1979) ; Engvall and Perlmann, J. Immunol.. 109: 129 (1972); Geoghegan et al., Immunol. Comm.. 7: 1 (1978) ; and Wilson and Nakane, Immunofluorescence and Related Technigues. p. 215 [Elεevier/North Holland Biomedical Press; Amsterdam (1978)]. Depending upon the nature of the label, various techniques can be employed for detecting, or detecting and quantitating the label. For fluorescers, a large number of fluorometers are available. For chemiluminescers, luminometerε or filmε are available. With enzymeε, a fluoreεcent, chemilumineεcent, or colored product can be determined or measured fluorometrically, luminometrically, spectrophotometrically or visually.

The following exampleε are preεented to help in the better understanding of the subject invention and are for purpoεeε of illuεtration only. The exampleε are not to be conεtrued as limiting the invention in any manner.

Example 1

In this example, one maternal urine sample from a Down εyndrome affected pregnancy at the geεtational age of 12 weekε, 0 dayε, as determined by ultrasound, was assayed for UGP according to the methods of thiε invention. [The case sample (karyotype: 47, XX, +21) waε kindly provided by Nancy C. Roεe, M.D. of the Department of Obεtetrics and Gynecology at the Hospital of the University of Pennsylvania (PA, USA) . ] The reason for the prenatal diagnosiε haε advanced maternal age. Six controls from gestational ages plus or minus one week from that of the Down syndrome case sample were tested by the same method to establiεh the MOM. [Five of the control εamples and one other εample were kindly provided reεpectively by Britta Stirnal, M.D. (Germany) , and Leonard H. Keliner, M.S. (Montefiore Medical Center, Albert Einεtein College of Medicine; New York, N.Y. USA) . The testε and analyses were performed by or under the direction of Leonard H. Kellner, M.S. (Montefiore Medical Center) and Jacob A. Canick, Ph.D. (Women and Infants Hospital, Brown University School of Medicine; Providence, Rhode Island, USA).]

Case and control samples were stored at -20°C for from 1 week to 8 months prior to asεay. While the fetal karyotype of the control εampleε waε not alwayε known, it waε aεsumed that none of the controls was from an aneuploid pregnancy. Gestational age of the control samples were determined by either ultrasound parameters or by last menstrual period dating.

The Triton" UGP EIA Kit from Ciba Corning Diagnostics (Alameda, CA; U.S.A.) was used to determine the UGP levels in the case and control sampleε. The protocol used for the assay was esεentially aε provided with the kit but appropriate dilutionε for the firεt trimester samples were made. The

assay range for UGP was 0.4-16.0 fmol/ml. Because of the high levels of UGP in the maternal urine samples, the samples prior to assay were diluted 50,000 to 100,000 (50k to 100k) with diluent provided with the kit as indicated in the results shown in Table 1.

UGP levels were ultimately expressed as pmol per mg urinary creatinine to account for variationε in the concentration of the urine samples. There had been no attempt to control the time of day that the urine sampleε were taken. Creatinine levelε were measured on a Synchron CX-5 chemistry analyser (Beckman Instrumentε; Brea, CA; USA) .

TABLE 1

UGP gest. creatinine UGP raw dilution UGP (pmol/mg

Samples age mg/ml (fmol/ml) factor (pmol/ml) creat.)

Control 1 13,2 1.01 2.34 100K 234.0 231.7

Control 2 11,6 0.63 1.48 50K 74.0 117.5

Control 3 12,1 0.52 2.54 50K 127.0 244.2

Control 4 11 ,4 1.10 4.05 50K 202.5 184.1

Control 5 11 ,0 0.90 3.06 50K 153.0 170.0

Control 6 12,0 1.04 0.83 50K 41.5 39.9

Case:

Down

Syndrome

Case 12,0 1.60 6.24 100K 624.0 390.0

The control values ranged from 39.9 to 244.2 pmol UGP per mg creatinine. The median value for the six controls was 177 pmol per mg creatinine. The level of UGP in the Down syndrome case was 390 pmol UGP per mg creatinine. Thus, the result for this case expressed as multiples of the control median (MOM) was 390/177 = 2.20 MOM, i.e., slightly more than twice the control median, a positive result.

Example 2 In this example, seven first trimester maternal urine samples from Down syndrome affected pregnancies, that is, seven case samples, and 214 first trimester control urine samples were assayed for UGP. The asεay protocol with the Triton UGP EIA Kit (CCD; Alameda, CA; USA) used was essentially as described above in Example 1. However, the samples were diluted 10,000-fold with the kit diluent, and if the UGP levels were εtill too high for the kit's assay range, a repeat assay was done at a higher dilution.

Dilutions were performed with a Hamilton Microlab 100 semi-automatic diluter. The concentrations reported in this example are in mmol UGP/L and expresεed aε nmol/mmol creatinine.

Controls

A regression curve from the UGP/creatinine values for the 214 control sampleε from pregnancies of 9 to 13 weeks gestation was prepared. The following quadratic equation basically describes the data: a + bx + cx ² + dx ³ wherein x represents the gestational age in days; and wherein the coefficients are as follows: a = -527.7; b = 19.13; c = -0.2230; and d = 0.0008558.

The observed versus the fitted values were as follows:

Gestational No. of

Week Samples Observed Fitted

9 12 9.9758 9.4130

10 18 12.6759 13.5389

11 77 14.5134 13.7500

12 29 11.4615 12.5007

13 78 10.9880 10.9107

There appeared to be a peak urinary UGP level at about 11 weekε and then a decline in the levelε thereafter.

MOMε in unaffected pregnancieε. The spread of the MOM valueε for unaffected pregnancies are exemplified as follows.

Controls UGP MOM 10th centile 0. 41

Median 0. 99 90th centile 1. 95

Assuming a Gausεian distribution, the SD of log UGP based on the 10th-90th range would be 0.26, which is similar to that of free beta hCG in serum.

Down Syndrome Affected Pregnancies

The resultε of teεting the seven urine sampleε from the Down εyndrome affected pregnancies are expressed below as multipleε of the geεtation-εpecific median (MOM) and shown below in Table 2. Case No. 3 is from a twin pregnancy discordant for Down syndrome.

TABLE 2

Down Svndrome Geεtational

Case No. Age in Weeks MOM

(1) 10 1.90215

(2) 11 0.34949

(3) (twin) 12 2.65054

(4) 12 0.92861

(5) 12 2.79836

(6) 12 0.49450

(7) 13 1.71604

Four out of the seven case samples extremely elevated, namely Case Noε. 1, 3, 5 and 7. The distribution in Down syndrome cases may be bimodal with the low values being

from non-viable pregnancies. About one-half of first trimester Down syndrome affected fetuses are non-viable.

Two of the urine samples — Case Nos. 3 and 7 — were taken after CVS. The remainder were prospective samples taken before CVS or any other invasive prenatal diagnostic procedure was performed.

The mean approximated the observed log median of the seven cases, the log of 1.72 MOM or 0.24. The observed SD of log UGP iε 0.36.

Covariables

There appears to be no correlation between the UGP concentrations in the first trimester urine εamples with the following covariables: maternal age (r = 0.04615 based on 72 samples); maternal weight (r = -0.21217 based on 65 samples); serum AFP (r = 0.02184 based on 67 sampleε); and εerum uE3 (r = 0.00305 based on 67 sampleε). However, there appears to be a statistically significant correlation between the urinary UGP levels and serum free beta hCG levelε, although the correlation coefficient is very low (r = 0.42888 based on 67 samples; p = 0.0003).

Example 3 Four hundred .and sixty-three sampleε were obtained from apparently normal pregnancieε at 9-13 weeks gestation. Five of those pregnancies were subsequently found to be affected with aneuploidy. An additional 7 samples were obtained after chorionic villus sampling (CVS) had confirmed aneuploidy. Of the 12 cases, 10 were affected by Down syndrome (trisomy 21) and 2 had Edwardε εyndrome (triεomy 18) . There were three sources of sampleε from apparently normal pregnancieε: 165 were obtained prior to CVS because of increased riεk of aneuploidy, 44 at the time of the firεt antenatal hoεpital viεit and 249 concurrently with maternal εerum εcreening which in Leedε [Centre for Reproduction, Growth & Development; Reεearch School of Medicine; Univerεity of Leedε (Leeds, U.K.)] is done at 13-18 weeks gestation although some women provide earlier sampleε in error.

Urine samples from cases and controls were stored at -20°C or lower until retrieved for assay. Each sample was assayed without knowledge of whether it was from a case or a control. UGP was meaεured using the enzyme immunoassay method [Triton" UGP EIA Ciba Corning Diagnostics Corp., Alameda, CA (USA) ] essentially as deεcribed above in Example 2. To allow for variable urine concentrationε, UGP levelε were expressed as nmol per mmol creatinine and to account for gestational differences, levels were alεo expreεεed aε multipleε of the normal median (MOM) level for the appropriate gestation. The normal gestation-specific median levels were derived by regresεion of the observed weekly medians against gestation among the controls after weighting for the number at each week. Gestational age was based on ultrasound biometry in each case and in 87% of controls.

The median UGP level increaεed to a peak at 10-11 weekε and fell thereafter. A log quadratic regression fitted thiε reaεonably well: the obεerved and regressed values were 10.9 and 12.2 nmol/mmol at 9 weeks (15 εamples); 13.3 and 13.2 at 10 weekε (58 εampleε); 14.6 and 13.5 at 11 weekε (85 samples); 10.5 and 13.0 at 12 weeks (38 samples); and 11.4 and 11.3 at 13 weeks (262 samples). The regresεion equation was log ₁₀ UGP = -1.079 + 0.05592*ga - 0.0003537*ga ² where ga is the gestation in days. Using this equation to calculate MOMs gave a normal median of 1.00, a 10-90th centile range of 0.40-2.08 and a 25-75th centile range of 0.61-1.51.

Table 3 εhowε the individual MOM values for each of the 12 affected pregnancies together with clinical detailε. The UGP levelε for both caεeε of Edwardε' εyndrome were below the normal 25th centile, a εtatistically εignificant reduction (p<0.05, Wilcoxon Rank Sum Teεt). In four of the 10 caεeε of Down syndrome the level was raised above the normal 90th centile, and in two it was reduced (one below the 10th centile and one below the 25th) . Overall the distributrion of valueε in Down εyndrome did not quite differ significantly from controls (P = 0.09, Wilcoxon Rank Sum Test).

Theεe reεultε indicate that firεt trimester maternal urine UGP levelε are raised in a proportion of pregnancies affected with Down syndrome and reduced in Edward syndrome.

Table 3

Individual Marker Level (MOM) in

12 Pregnancies with Aneuploidy

Aneuploidy MOM Geεtation Maternal Source (wkε) Age at Term (yrs)

Down 3.57 13 28 Post CVS

Down 3.27 11 45 Post CVS

Down 2.64 12 37 Post CVS

Down twin 2.53* 12 26 Post CVS

Down 1.37 10 32 Pre CVS

Down 1.06 11 28 Post CVS

Down 0.94 10 36 Pre CVS

Down 0.89 12 33 Pre CVS

Edwards 0.57* 12 40 Post CVS

Down 0.48 12 38 Pre CVS

Down 0.36* 11 37 Pre CVS

Edwards 0.10* 11 40 Post CVS

* The εampleε marked with an asterisk were also included in experiments using a different aεεay for UGP and yielded levels of 0.48, 0.48, 1.01 and 0.06 MOM reεpectively. The aεεay uεed waε a modification of a radioimmunoaεsay method described in Lee et al., J. Endocrinol.. 13: 481-489 (1991) as described in Cuckle et al., Prenatal Diagnoεis. 14: 953- 958 (1994).

Example 4 Multiple Marker Urinary Screening — One First Trimester Down Syndrome Case and One First Trimester Turner Syndrome Case. Tested for UGP and Free α-hCG by RIA. and tE by Kober Reaction

In this example, UGP levels were detected and quantitated in urine along with two other markers — tE and free α-subunit of hCG (α-hCG) . First trimester maternal urine samples were tested from two pregnancies affected by aneuploidy — one with Down syndrome (trisomy 21) and one with Turner syndrome (45,X/46,X+ring — a mosaic) . Maternal urine samples were also taken from 291 controls, of which 49 were at a gestational age of <13 weeks, and 77 were at a geεtational age of 13-14 weekε.

Methods

In the case of the Down syndrome affected pregnancy, the urine sample was obtained at 11 weeks gestational age before prenatal diagnosis by chorionic villus sampling (CVS) . CVS was performed because of family history or chromosomal abnormality. In the case of the Turner syndrome affected pregnancy, the urine sample was taken at 13 weeks gestational age after CVS that was performed becauεe of advanced maternal age.

Patients were recruited in the area served by the Yorkshire Regional Cytogenetics Laboratory [Yorkshire, UK] and from those attending selected antenatal clinics at Queen Charlotte's Maternity Hospital [London, UK]. A control εeries was obtained from the urine samples which are routinely taken for sugar, protein and bacterial analysis from women attending antenatal clinics in the Leeds General Infirmary [Leeds, UK] and at Queen Charlotte's Hospital.

Urine samples from caseε and controls were stored at -40°C until retrieved for aεεay. Each sample was assayed without knowledge of whether it was from a case or a control. UGP was measured using a modification of a previously published radioimmunoaεεay method [Lee et al., J. Endocrinol..

130: 481-489 (1991)]. The UGP specific antisera (S504- Immunogen International, Llandyssul, Dyfed SA44 5JT, UK) was retitrated against UGP tracer in order to give a desensitized asεay range of 6-600 pmol/L. Thiε aεsay has a partial mole per mole crosε-reaction with intact hCG (6.9%) and free β-hCG (18%) but neglible croεε-reactivity with free α-εubunit hCG, luteinizing hormone (LH) , thyroid stimulating hormone (TSH) and follicle εtimulating hormone (FSH) (<0.7%). All εamples were diluted 1/100 and 1/1000 prior to aεεay. tE waε measured by a continuous flow system based on the Kober reaction [Lever et al., Biochem. Med.. 8: 188-198 (1973)], senεitized by doubling the sample volume and having the concentration of the calibrators previously used for late pregnancy samples. Free α-hCG was measured using an in-house desenεitized radioimmunoaεsay with a standard range of 25-500 μg/L, thus requiring εamples to be diluted 1/5 or 1/10. The assay uses a polyclonal sheep antibody raised against free α- hCG isolated from a crude commercial preparation of pregnancy urine hCG by gel exclusion and hydrophobic interaction chromatography. An immunoglobulin preparation from the resultant antisera (S781; Polyclonal Antibodies Ltd, Llandyssul, Dyffed, UK) was extensively absorbed against a solid phase of intact hCG conjugated to activated sepharose. This removed antibodieε to epitopes present on free α-hCG expoεed when it combineε with the β-subunit. The NIH reference preparation CR123 was used to prepare assay standards and to form the tracer, by iodination using the chloramine T method. Croεs-reactivity with intact hCG, free β-hCG, UGP, luteinizing hormone (LH) , follicle stimulating hormone (FSH) and thyrotropin (TSH) was asseεεed at 50% of B/B _Q . Purified reference preparations were obtained from the National Institutes of Health (Bethesda, Maryland, USA) for intact hCG (CR123) , free β-hCG (CR123) and UGP [donated by Drs. R. Wehmann and D. Blithe; Blithe et al., Endocrinol.. 122: 173-180 (1988)] and WHO International Reference Preparations from the National Institute for Biological Standards and Control (Potterε Bar, Hertfordεhire, UK) for LH

(68/40) , FSH (83/575) and TSH (80/558) . Molecular weights were calculated from the established primary structures: intact hCG 38 kd; free β-hCG 23 kd; free α-hCG 15 kd [Bellisario et al., J. Biol. Chem.. 248: 6796 (1973)]; UGP 10 kd [Birken et al., Endocrinol.. 123: 572 (1988)]; LH 29 kd; FSH 33 kd and TSH 30 kd (Professor S. Jeffcoate, personal communication) . On a molar basiε the cross-reactivity was <0.01% with free β-hCG and UGP but higher with intact hCG (10%) , LH (8%) , FSH (8%) and TSH (6%) , possibly caused by dissociation of the purified preparations which would free the α-subunit common to all the glycoprotein hormones.

Since random urines were obtained and there was no uniformity in the time of voiding, the creatinine concentration was measured (Jaffe method using a Monach 200 centrifugal analyεer) , and the concentrationε of the markerε were expreεsed per mmol creatinine. To allow for gestational differences, levels were also expressed as multiples of the normal median (MOM) level for the appropriate gestation derived from the 291 singleton controls. Becauεe of the small number of controls at any week of geεtation, and εo that MOMε could be calculated to the exact day of geεtation, the obεerved medians in different gestational bands were regresεed after weighting for the number in each band. Gestational age was based on ultrasound biparietal diameter or crown-rump length measurement for all caseε and controlε.

Standard statistical modeling techniques were used to examine the screening potential of each marker and of different multi-marker combinations [Royston and Thompson, Stats. Med.. 11: 257 (1992)]. A multivariate log Gaussian frequency distribution was assumed, provided no significant deviation from fit was found in the Shapiro-Wilks Test. If the fit was poor in the extreme tails of the distribution, truncation limits were used. The marker means were estimated by the observed medianε, the εtandard deviationε by the 10- 90th centile differenceε divided by 2.563 and the correlation coefficients by the observed values after excluding outliers exceeding 3 standard deviations from the mean. The expected Down syndrome detection rate was calculated for a given false-

positive rate from the log Gaussian model, assuming that the maternal age distribution is that of England and Wales in 1989-92 (Office of Population Censuseε & Surveys (1991-1994) .

Results The observed median level of each marker in unaffected singleton pregnancies according to gestational age are shown in Table 4. The regressed medians used to calculate M _O Ms were IO ² - ⁸⁷¹ -°- ⁰¹ 505XGA _{for υGP f 10} -I.271+O.OI31XGA _{for tE and 10} -

0.170 ⁺ 0.001 ⁸ XGA _{for α} _ _{hCG/ w} ere _G A is the gestation in _d ays. Table 5 shows the individual MOM values for each of the two aneuploidy affected pregnancies together with the clinical details. For comparison. Table 6 shows selected centiles in the 291 singleton controls. There was a wide spread of UGP values in the εingleton controls exemplified by the 3-fold range between the 25th and 75th centiles or by the 7-fold 10th to 90th centile range.

TABLE 4 Median Marker Level in Singleton Unaffected Pregnancies According to Gestational Age

Gestation Pregnancies UGP tE α- -hCG

(wks) nmol/1 nmol/mmol μmol/l μmol/mmol nmol/1 nmol/mmol creatinine creatinine creatinine

< 13 49 273 33.7 7.0 0.71 9.7 1.06

13-14 77 235 23.6 11.0 1.12 12.4 1.06

TABLE 5 Individual Marker Level (MOM) in 1 Singleton I Down Syndrome Pregnancy and 1 Turner Syndrome Pregnancy I

Aneuploidy Gestation Maternal (wks) age (yrs) Diagnosiε* UGP tE α-hCG

Down syndrome 11 37 FH → CVS ⁺ 1.01 0.54 1.09 Turner syndrome 13 42 MA → CVS 3.35 0.37 0.60

Reasons for carrying out prenatal diagnosis → method used: FH = family history or chromosomal abnormality; MA = advanced maternal age; CVS = chorionic villus sampling.

Urine sample obtained before the prenatal diagnosis.

TABLE 6 Selected Centileε of Marker Level (MOM) in 291 Singleton Unaffected Pregnancies

Centile UGP tE α-hCG

10th 0.38 0.60 0.54

25th 0.66 0.80 0.68

50th 1.04 0.99 1.00

75th 1.66 1.35 1.61

90th 2.50 1.77 3.22

The resultε εhown in Table 5 indicate a false- negative result for UGP level in the Down syndrome case, but the level of UGP for the Turner syndrome case, more than three times above normal, indicateε a poεitive reεult. The tE level in the Down syndrome case is a positive result being about one-half of the normal level. The tE level in the Turner εyndrome case being at 0.37 MOM is also indicative of a positive result. The α-hCG level for the Down syndrome caεe iε a falεe-negative result at 1.09 MOM, whereas the α-hCG result for the Turner syndrome case is a poεitive reεult aε 0.60 MOM iε below normal levels.

The descriptionε of the foregoing embodiments of the invention have been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise form discloεed, and obviously many modifications and variations are possible in light of the above teachings. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application to enable thereby others skilled in the art to best utilize the invention in various embodiments and with various modificationε as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims appended hereto. All references cited herein are hereby incorporated by reference.