American Journal of Medical Genetics 3 7 114-1 18 (1990)
Second Trimester Maternal Serum Pregnancy Specific Beta-1 Glycoprotein (SP-1)Levels in Normal and Down Syndrome Pregnancies Elizabeth Petrocik, E. Robert Wassman, J. Jack Lee, and JoAnn C. Kelly ALFIGENIThe Genetics Institute, Pasadena (E.P., E.R.W., J.C.K.); Department of Bwstatistics, School of Public Health, University of California at Los Angeles, (J.J.L.) California Maternal serum pregnancy specific beta-1 glycoprotein (SP-1) levels in the second trimester may be predictive of Down syndrome (DS). An enzyme immunoassay was used to measure SP-1 sera from 46 DS pregnancies and 117 normal control women matched for maternal age, gestational age, and length of storage. In the normal control samples, there were slight correlations between the SP-1concentration and maternal age. The maternal serum SP-1levels increased with each week of gestation from 15 to 20 weeks. All but one of the DS sera had SP-1 levels greater than the normal median. Using a cutoff of 2.8 multiples of the median (MOM),15.2%of the DS pregnancies were detected with a false-positiverate of 4.3%. A combinational logistic regression analysis of maternal age and pregnancy related serum proteins will detect additional DS pregnancies and decrease the false-positive rate. The combination of maternal age and SP-1 detected 33 (71.7%)of Down syndrome pregnancies. The addition of maternal serum a-fetoprotein (AFP)and human chorionic gonadotropin (hCG) levels allowed for the detection of 36 (78.3%) of the DS pregnancies with a decrease in the false-positive rate to 3.4%. The measurement of other serum constituents in conjunction with AFP appears to be a valuable addition to current screening programs, as this can increase the proportion of DS cases detected prenatally.
KEY WORDS: Down syndrome, SP-1,logistic regression, pregnancy
Received for publication October 23,1989; revision received January 22, 1990. Address reprint requests to Dr. JoAnn C. Kelly, ALFIGEN/The Genetics Institute, 11 W. Del Mar Blvd., Pasadena, CA 91105.
0 1990 Wiley-Liss, Inc.
INTRODUCTION Down syndrome (DS),the most common form of inherited mental retardation, occurs in approximately 1 of 800 live births [Goodwin and Heuther, 19871. Since advanced maternal age is strongly correlated with DS, the current standard of care is t o offer diagnostic amniocentesis to all pregnant women over age 35 years. In addition, low maternal serum a-fetoprotein (MSAFP) concentrations have been associated with an increased rate offetuses with trisomy 21 [Merkatz et al., 19841. Consequently, at many centers, amniocentesis is offered when the combination of low MSAFP concentrations and maternal age produces a mid-trimester risk of greater than approximately 1: 270 [Cuckle et al., 1984; Doran et al., 19861. This allows prenatal detection of about 20-30% of all DS pregnancies. Other maternal serum markers may also be predictive of DS pregnancies. Low second trimester maternal serum unconjugated estriol has been associated with an increased risk of a DS pregnancy [Canick et al., 1988; Wald et al., 19881. Several recent studies have documented a strong association between elevated maternal serum human chorionic gonadotropin (hCG) concentrations and the risk of DS [Bogart et al., 1987; Petrocik et al., 19891.Wald et al. [1989bl were able to detect 61% of DS pregnancies using unconjugated estriol, hCG, MSAFP, and maternal age. We have been able to duplicate this level of detection using only MSAFP, hCG, and maternal age, but using a unique logistic regression analysis [Petrocik et al., 19891. A recent retrospective study of Bartels and Lindemann [19881suggested that maternal serum pregnancy specific beta-1 glycoprotein (SP-1) concentrations in mid-trimester may be used to detect DS pregnancies. In their study, all but one of 24 DS pregnancies had SP-1 levels above the medians for normal pregnancies. Wald et al. [1989al found only 31% of DS pregnancies had SP-1 levels above 2.8 MOM.The study presented here was designed to investigate further those observations and differences. Maternal serum SP-1 concentrations were measured in the second trimester of DS and matched control pregnancies. Sera from normal pregnancies were analyzed to determine which, if any, clinical criteria, including gestational age, maternal age,
SP-1 Levels in Down Syndrome Pregnancies
115
and weight, should be corrected for during evaluation of the data. In addition, we designed a method for combinational calculation of risks based on MSAFP, hCG, and SP-1values. These studies suggest that SP-1 may be a valuable addition to MSAFP and hCG in the detection of DS pregnancies.
tional age, and storage time were thawed and assayed for all 3 proteins. All gestational ages were calculated from the biparietal diameter measurement as determined by ultrasound examination or, when not available, from the first day of the last menstrual period. Statistical methods are described in Mendenhall [ 19791.
MATERIAL AND METHODS Serum SP-1concentrations were determined using an enzyme immunoassay (EIA),Enzygnost-SP1, manufactured by Behringwerke AG, Marburg, FRG. Permission was obtained from the Federal Department of Agriculture and the U S . Department of Agriculture to import and employ this EIA for research purposes only. The assay was modified only in that the experiment was carried out in a dry thermostated incubator rather than in a water bath. The results are reported in international units per milliliter. Serum p-hCG and AFP levels were measured using EIA kits produced by Abbott Diagnostics laboratories. Standards employed in the p-hCG kit were calibrated against the World Health Organization (WHO) Second International Standard for hCG. Dilution to 1:500 or 1 : 1,000 was the only alteration made in the P-hCG protocol. Variation within each hCG assay was calculated to be 4.8% while variation between runs was determined to be 6.1%. The results are reported in international units per milliliter. The AFP-EIA standards were calibrated using the WHO First International Standard of AFP. The coefficient of variation for AFP was 5.8%within runs and 6.9%between runs. The AFP results are reported in nanograms per ml. Serum specimens were collected prior t o amniocentesis from 46 DS pregnancies between January 1,1987 and July 31, 1989 and stored frozen at -20°C until analyzed for P-hCG, AFP, and SP-1 concentrations. Indications for amniocentesis were advanced maternal age (n = 39), low MSAFP (n = 4), and abnormal ultrasound (n = 3). All diagnoses were confirmed after termination or delivery of a DS baby; 117 randomly selected control samples from women matched for maternal age, gesta-
RESULTS In the normal population, maternal serum SP-1concentrations showed a slight negative correlation with maternal age, r = -0.21, P = 0.03 (Fig. 1).Figure 2 shows the independence of maternal weight and serum SP-1 levels (r = -0.07, p = 0.48) in normal pregnancies. No corrections for maternal age or weight were made in this study. SP-1 levels increased linearly with each completed week of gestation from 65 IU/ml at 15 weeks to 111 IU/ml at 19 weeks. Maternal serum SP-1 levels were compared with MSAFP and hCG levels in the normal controls. It can be seen from Figure 3 that serum AFP and SP-1 are not correlated (r = -0.003, P = 0.97). Neither are serum SP-1concentrations and serum hCG levels closely correlated (r = 0.12, P = 0.20), as shown in Figure 4. The independent variation of SP-1levels from maternal age, weight, hCG and AFP hold within as well as across gestational age (data not shown). Medians were generated from the control population for 15-20 weeks gestation. The DS pregnancy serum SP-1 values were plotted by gestational week in Figure 5. All but one of the DS serum SP-1 levels were above the normal median. Using maternal serum SP-1 levels alone with a cutoff of 2.8 MOM, 15.2% of the DS cases were detected with a false-positive rate of 4.3% (TableI). Lowering the cutoff to 2.0 MOMdetected more of the DS pregnancies (43%)but increased the false-positive rate to 13.6%.The DS detection rate with SP-1 alone (15.2%) was comparable to that of hCG alone (23.9%)and AFP alone (15.2% using a cutoff of 0.6 MOM). The SP-1 MOMSand maternal ages of the 46 DS pregnancies and the 117 normal pregnancies were entered into a stepwise logistic regression analysis in order t o
. SP-1
I
IS
2o
2s
30 AGE (YEARS)
3s
40
15
Fig. 1. Maternal serum SP-1 concentration (IU/ml) by maternal age (in years) in normal pregnancies.
116
Petrocik et al. 450
T
8
8
80
100
140
120
160
200
180
220
240
WEIGHT (LBS.) Fig. 2. Maternal serum SP-1 concentration (IUlml) by weight (in pounds) in normal pregnancies.
ery, and, with a half-life of 30 hr, is no longer detectable 2 weeks postpartum [Dati et al., 19821. This study and those of Bartels and Lindemann [19881 and Bartels et al. [19901 evaluated SP-1 levels in patients during the second trimester, when maternal serum concentrations are slowly rising. This study was done to evaluate the feasibility of the use of SP-1 screening in conjunction with MSAFP (and hCG) screening programs. Interpretation of MSAFP and hCG values is based on MOMfor each week of gestation [Milunsky et al., 1988; Bogart et al., 1987; Petrocik et al., 19891. For maternal serum SP-1, the observed DISCUSSION values increased with each week of gestation and were SP-1is synthesized by cells in the syncytiotrophoblast consistent enough to permit the determination of pre[Horne et al., 19761. It is secreted into the maternal liminary medians between 15and 20 weeks of gestation. blood and can be detected as early as 21 days after the This suggests that maternal serum SP-1 levels can be last menstrual period (LMP) [Anthony et al., 1980; interpreted on the basis of MOM. Interpretation of maternal serum SP-1concentration Grudzinskas et al., 1977; Eiermann et al., 19811. The SP-1 concentration in maternal serum increases does not appear to require corrections for maternal age throughout pregnancy, diminishes rapidly after deliv- or weight. Since SP-1 is produced by the placenta, fac-
generate an equation that would predict a woman's risk of bearing a DS fetus [Aldrich and Nelson, 19841. This form of combinational analysis calculated each woman's probability of carrying a DS baby that was 71.7% predictive of trisomy 21 with a 4.3% false-positive rate. Analysis of the same data set on a probit regression model was equally predictive of DS. The detection rate was slightly better with the addition of MSAFP-MOMSand hCG-MoMs, which allowed for the detection of 36 (78.3%) of the DS pregnancies (3 additional) and a decrease in the false-positive rate to 3.4%.
8
8
- 0 0
2o
40
,
I
60
80
100
8
8
120
140
AFP Fig. 3. Maternal serum SP-1 concentration (IU/ml) versus MSAFP (in ng/ml) in normal pregnancies
450
.
-
"T m
SP-1 Levels in Down Syndrome Pregnancies
117
8
SP-1
200 150 100
50
0
c 0
10
30 HCQ (IUML)
u)
40
50
60
Fig. 4. Maternal serum SP-1 concentration (IU/ml) versus maternal serum hCG (in IU/ml) in normal pregnancies.
300 250
c '-
ID
S P - 1 200 .150 -
IW-50 .-
. . . . .. . . . . . . '. ... . 8
I
.
I.
.
==
m
8
8 m
1'
9
.I
8
8
0 -
tors that influence placental size may alter the maternal serum SP-1levels. Therefore, further studies are needed to determine the effect of smoking, maternal diabetes, and multiple gestation. By measurement of maternal serum SP-1 levels alone, 15.2%of the DS pregnancies were detected with a falsepositive rate of 4.3%.This level of detection, which was lower than that observed by Bartels and Lindermann [19881 and by Wald et al. 119891, was comparable to maternal serum hCG alone and MSAFP alone. Combining maternal serum SP-1concentrations with maternal serum AFP, hCG, and maternal age a t the time of delivery, through a logistic multiple regression analysis, it was possible to formulate a risk factor for each patient that was 78.3% predictive of DS with a false-positive rate of 3.4%.Without AFP and hCG included in this model, 3 less pregnancies were detected (Table I). Undoubtedly, inherent biases exist in the present study that may mitigate the predictiveness of this method of combinational analysis. While the model may hold great value for screening programs in the future,
I
the lack of readily availabile assay kits or procedures may limit its present application. Additional studies, including prospective studies, are necessary for adequate assessment of its value to prenatal screening for DS. TABLE I. Number and Proportion of Predicted DS Pregnancies* Down syndrome pregnancies n %
Normal pregnancies n %
AFP only (MOM 2.9) SP-1 only (MOM >2.8) Combined maternal age and SP-1 Combined maternal age, AFP, hCG, and SP-1 *Normal pregnancies n
=
7 5 5 5
4.3 4.3 4.3
7 11 7 33
15.2 23.9 15.2 71.7
4
3.4
36
78.3
6.0
117; DS pregnancies n
=
46.
118
Petrocik et al.
REFERENCES Aldrich JH, Nelson FD (1984): “Linear Probability Logit and Probit Models.” Beverly Hills, California: Sage Publications. Anthony R, Masson GM, Wood PJ (1980): The radioimmunoassay of pregnancy specific B1-glycoprotein in plasma as a pregnancy test for subfertile women. Br J Obstet Gynaecol 87:496-500. Bartels I, Lindemann A (1988): Maternal levels of pregnancy-specific B1-glycoprotein (SP-1) are elevated in pregnancies affected by Down’s syndrome. Hum Genet 80:46-48. Bartels I, Thick M, Bogart MH (1990): Maternal serum hCG and SP1in pregnancies with fetal aneuploidy. Am J Med Genet (in press). Bogart MH, Pandian MR, Jones OW (1987): Abnormal maternal serum chorionic gonadotropin levels in pregnancies with fetal chromosome abnormalities. Prenat Diagn 7:623-630. Bogart MH, Golbus MS, Sorg ND, Jones OW (1989): Human chorionic gonadotropin levels in pregnancies with aneuploid fetuses. Prenat Diagn 9:379-384. Canick JA, Knight GJ, Palomaki GE, Haddow JE, Cuckle HS, Wald NJ (1988):Low second trimester maternal serum unconjugated oestriol in pregnancies with Down’s syndrome. Br J Obstet Gynaecol 95:330-333. Cuckle HS, Wald NJ, Lindebaum RH (1984): Maternal serum alpha fetoprotein measurement: A screening test for Down syndrome. Lancet 1:926-929. Dati F, Grenner G, Luben G, Kapmeyer W, Sieber A, Bohn H, Bellmann 0 (1982):Comparison of enzyme-immunoassays for SP1 and AFP with other immunological methods. La Ricerca Clin Lab. Doran TA, Cadesky K, Wong PY, Mastrogiacomo C, Capello T (1986): Maternal serum alpha fetoprotein and fetal autosomal trisomies. Am J Obstet Gynecol 154:277-281. Eiermann W, Albrich W, Dati F, Leis D, Eicher W (1981): Enzymimmunoassay: Klinische Anwendung. I. Schwangerschaftsfruhdiagnose. Geburtsh Frauenheilkd 41:404-406.
Goodwin BA, Heuther CA (1987):Revised estimates and projections of the Down syndrome births in the United States, and the effects of prenatal diagnosis utilization, 1970-2002. Prenat Diagn 7:261-271. Grudzinskas G, Gordon YB, Jeffrey D, Chard T (1977): Specific and sensitive determination of pregnancy-specific B1-glycoprotein by radioimmunoassay. Lancet 1:331-335. Horne CHW, Towler CM, Pugh-Humpreys RGP (1976):Pregnancy specific B1-glycoprotein-A product of the syncytiotrophoblast. Experientia 32:1197-1201. Mendenhall W (1979): “Introduction to Probability and Statistics.” 5th Ed. North Scituate, Massachusetts: Duxbury Press. Merkatz IR, Nitowsky HM, Marci JN, Johnson WE (1984): An association between low maternal serum alpha fetoprotein and fetal chromosome abnormalities. Am J Obstet Gynecol 148:886-894. Milunsky A, Wands J , Brambati B, Bonacchi I, Currie K (1988): First trimester maternal serum fetoprotein screening for chromosome defects. Am J Obstet Gynecol 159:209-213. Petrocik E, Wassman ER, Kelly J C (1989): Prenatal screening for Down syndrome with maternal serum human chorionic gonadotropin levels. Am J Obstet Gynecol 161:1168-1173. Wald N, Cuckle H, Densem J (19891: Maternal serum specific beta 1-glycoprotein in pregnancies associated with Down’s syndrome. Lancet 2:450. Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Royston P, Chard T, Haddow J E , Knight GJ,Palomaki GE, Canick JA (1989): Maternal serum screening for Down’s syndrome in early pregnancy. Br Med J 297:883-887. Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Canick JA, Haddow JE, Knight GJ, Palomaki GE (1989): Maternal serum unconjugated oestriol as an antenatal screening test for Down’s syndrome. Br J Obstet Gynaecol 95:334-341.
Second Trimester Maternal Serum Pregnancy Specific Beta-1 Glycoprotein (SP-1)Levels in Normal and Down Syndrome Pregnancies Elizabeth Petrocik, E. Robert Wassman, J. Jack Lee, and JoAnn C. Kelly ALFIGENIThe Genetics Institute, Pasadena (E.P., E.R.W., J.C.K.); Department of Bwstatistics, School of Public Health, University of California at Los Angeles, (J.J.L.) California Maternal serum pregnancy specific beta-1 glycoprotein (SP-1) levels in the second trimester may be predictive of Down syndrome (DS). An enzyme immunoassay was used to measure SP-1 sera from 46 DS pregnancies and 117 normal control women matched for maternal age, gestational age, and length of storage. In the normal control samples, there were slight correlations between the SP-1concentration and maternal age. The maternal serum SP-1levels increased with each week of gestation from 15 to 20 weeks. All but one of the DS sera had SP-1 levels greater than the normal median. Using a cutoff of 2.8 multiples of the median (MOM),15.2%of the DS pregnancies were detected with a false-positiverate of 4.3%. A combinational logistic regression analysis of maternal age and pregnancy related serum proteins will detect additional DS pregnancies and decrease the false-positive rate. The combination of maternal age and SP-1 detected 33 (71.7%)of Down syndrome pregnancies. The addition of maternal serum a-fetoprotein (AFP)and human chorionic gonadotropin (hCG) levels allowed for the detection of 36 (78.3%) of the DS pregnancies with a decrease in the false-positive rate to 3.4%. The measurement of other serum constituents in conjunction with AFP appears to be a valuable addition to current screening programs, as this can increase the proportion of DS cases detected prenatally.
KEY WORDS: Down syndrome, SP-1,logistic regression, pregnancy
Received for publication October 23,1989; revision received January 22, 1990. Address reprint requests to Dr. JoAnn C. Kelly, ALFIGEN/The Genetics Institute, 11 W. Del Mar Blvd., Pasadena, CA 91105.
0 1990 Wiley-Liss, Inc.
INTRODUCTION Down syndrome (DS),the most common form of inherited mental retardation, occurs in approximately 1 of 800 live births [Goodwin and Heuther, 19871. Since advanced maternal age is strongly correlated with DS, the current standard of care is t o offer diagnostic amniocentesis to all pregnant women over age 35 years. In addition, low maternal serum a-fetoprotein (MSAFP) concentrations have been associated with an increased rate offetuses with trisomy 21 [Merkatz et al., 19841. Consequently, at many centers, amniocentesis is offered when the combination of low MSAFP concentrations and maternal age produces a mid-trimester risk of greater than approximately 1: 270 [Cuckle et al., 1984; Doran et al., 19861. This allows prenatal detection of about 20-30% of all DS pregnancies. Other maternal serum markers may also be predictive of DS pregnancies. Low second trimester maternal serum unconjugated estriol has been associated with an increased risk of a DS pregnancy [Canick et al., 1988; Wald et al., 19881. Several recent studies have documented a strong association between elevated maternal serum human chorionic gonadotropin (hCG) concentrations and the risk of DS [Bogart et al., 1987; Petrocik et al., 19891.Wald et al. [1989bl were able to detect 61% of DS pregnancies using unconjugated estriol, hCG, MSAFP, and maternal age. We have been able to duplicate this level of detection using only MSAFP, hCG, and maternal age, but using a unique logistic regression analysis [Petrocik et al., 19891. A recent retrospective study of Bartels and Lindemann [19881suggested that maternal serum pregnancy specific beta-1 glycoprotein (SP-1) concentrations in mid-trimester may be used to detect DS pregnancies. In their study, all but one of 24 DS pregnancies had SP-1 levels above the medians for normal pregnancies. Wald et al. [1989al found only 31% of DS pregnancies had SP-1 levels above 2.8 MOM.The study presented here was designed to investigate further those observations and differences. Maternal serum SP-1 concentrations were measured in the second trimester of DS and matched control pregnancies. Sera from normal pregnancies were analyzed to determine which, if any, clinical criteria, including gestational age, maternal age,
SP-1 Levels in Down Syndrome Pregnancies
115
and weight, should be corrected for during evaluation of the data. In addition, we designed a method for combinational calculation of risks based on MSAFP, hCG, and SP-1values. These studies suggest that SP-1 may be a valuable addition to MSAFP and hCG in the detection of DS pregnancies.
tional age, and storage time were thawed and assayed for all 3 proteins. All gestational ages were calculated from the biparietal diameter measurement as determined by ultrasound examination or, when not available, from the first day of the last menstrual period. Statistical methods are described in Mendenhall [ 19791.
MATERIAL AND METHODS Serum SP-1concentrations were determined using an enzyme immunoassay (EIA),Enzygnost-SP1, manufactured by Behringwerke AG, Marburg, FRG. Permission was obtained from the Federal Department of Agriculture and the U S . Department of Agriculture to import and employ this EIA for research purposes only. The assay was modified only in that the experiment was carried out in a dry thermostated incubator rather than in a water bath. The results are reported in international units per milliliter. Serum p-hCG and AFP levels were measured using EIA kits produced by Abbott Diagnostics laboratories. Standards employed in the p-hCG kit were calibrated against the World Health Organization (WHO) Second International Standard for hCG. Dilution to 1:500 or 1 : 1,000 was the only alteration made in the P-hCG protocol. Variation within each hCG assay was calculated to be 4.8% while variation between runs was determined to be 6.1%. The results are reported in international units per milliliter. The AFP-EIA standards were calibrated using the WHO First International Standard of AFP. The coefficient of variation for AFP was 5.8%within runs and 6.9%between runs. The AFP results are reported in nanograms per ml. Serum specimens were collected prior t o amniocentesis from 46 DS pregnancies between January 1,1987 and July 31, 1989 and stored frozen at -20°C until analyzed for P-hCG, AFP, and SP-1 concentrations. Indications for amniocentesis were advanced maternal age (n = 39), low MSAFP (n = 4), and abnormal ultrasound (n = 3). All diagnoses were confirmed after termination or delivery of a DS baby; 117 randomly selected control samples from women matched for maternal age, gesta-
RESULTS In the normal population, maternal serum SP-1concentrations showed a slight negative correlation with maternal age, r = -0.21, P = 0.03 (Fig. 1).Figure 2 shows the independence of maternal weight and serum SP-1 levels (r = -0.07, p = 0.48) in normal pregnancies. No corrections for maternal age or weight were made in this study. SP-1 levels increased linearly with each completed week of gestation from 65 IU/ml at 15 weeks to 111 IU/ml at 19 weeks. Maternal serum SP-1 levels were compared with MSAFP and hCG levels in the normal controls. It can be seen from Figure 3 that serum AFP and SP-1 are not correlated (r = -0.003, P = 0.97). Neither are serum SP-1concentrations and serum hCG levels closely correlated (r = 0.12, P = 0.20), as shown in Figure 4. The independent variation of SP-1levels from maternal age, weight, hCG and AFP hold within as well as across gestational age (data not shown). Medians were generated from the control population for 15-20 weeks gestation. The DS pregnancy serum SP-1 values were plotted by gestational week in Figure 5. All but one of the DS serum SP-1 levels were above the normal median. Using maternal serum SP-1 levels alone with a cutoff of 2.8 MOM, 15.2% of the DS cases were detected with a false-positive rate of 4.3% (TableI). Lowering the cutoff to 2.0 MOMdetected more of the DS pregnancies (43%)but increased the false-positive rate to 13.6%.The DS detection rate with SP-1 alone (15.2%) was comparable to that of hCG alone (23.9%)and AFP alone (15.2% using a cutoff of 0.6 MOM). The SP-1 MOMSand maternal ages of the 46 DS pregnancies and the 117 normal pregnancies were entered into a stepwise logistic regression analysis in order t o
. SP-1
I
IS
2o
2s
30 AGE (YEARS)
3s
40
15
Fig. 1. Maternal serum SP-1 concentration (IU/ml) by maternal age (in years) in normal pregnancies.
116
Petrocik et al. 450
T
8
8
80
100
140
120
160
200
180
220
240
WEIGHT (LBS.) Fig. 2. Maternal serum SP-1 concentration (IUlml) by weight (in pounds) in normal pregnancies.
ery, and, with a half-life of 30 hr, is no longer detectable 2 weeks postpartum [Dati et al., 19821. This study and those of Bartels and Lindemann [19881 and Bartels et al. [19901 evaluated SP-1 levels in patients during the second trimester, when maternal serum concentrations are slowly rising. This study was done to evaluate the feasibility of the use of SP-1 screening in conjunction with MSAFP (and hCG) screening programs. Interpretation of MSAFP and hCG values is based on MOMfor each week of gestation [Milunsky et al., 1988; Bogart et al., 1987; Petrocik et al., 19891. For maternal serum SP-1, the observed DISCUSSION values increased with each week of gestation and were SP-1is synthesized by cells in the syncytiotrophoblast consistent enough to permit the determination of pre[Horne et al., 19761. It is secreted into the maternal liminary medians between 15and 20 weeks of gestation. blood and can be detected as early as 21 days after the This suggests that maternal serum SP-1 levels can be last menstrual period (LMP) [Anthony et al., 1980; interpreted on the basis of MOM. Interpretation of maternal serum SP-1concentration Grudzinskas et al., 1977; Eiermann et al., 19811. The SP-1 concentration in maternal serum increases does not appear to require corrections for maternal age throughout pregnancy, diminishes rapidly after deliv- or weight. Since SP-1 is produced by the placenta, fac-
generate an equation that would predict a woman's risk of bearing a DS fetus [Aldrich and Nelson, 19841. This form of combinational analysis calculated each woman's probability of carrying a DS baby that was 71.7% predictive of trisomy 21 with a 4.3% false-positive rate. Analysis of the same data set on a probit regression model was equally predictive of DS. The detection rate was slightly better with the addition of MSAFP-MOMSand hCG-MoMs, which allowed for the detection of 36 (78.3%) of the DS pregnancies (3 additional) and a decrease in the false-positive rate to 3.4%.
8
8
- 0 0
2o
40
,
I
60
80
100
8
8
120
140
AFP Fig. 3. Maternal serum SP-1 concentration (IU/ml) versus MSAFP (in ng/ml) in normal pregnancies
450
.
-
"T m
SP-1 Levels in Down Syndrome Pregnancies
117
8
SP-1
200 150 100
50
0
c 0
10
30 HCQ (IUML)
u)
40
50
60
Fig. 4. Maternal serum SP-1 concentration (IU/ml) versus maternal serum hCG (in IU/ml) in normal pregnancies.
300 250
c '-
ID
S P - 1 200 .150 -
IW-50 .-
. . . . .. . . . . . . '. ... . 8
I
.
I.
.
==
m
8
8 m
1'
9
.I
8
8
0 -
tors that influence placental size may alter the maternal serum SP-1levels. Therefore, further studies are needed to determine the effect of smoking, maternal diabetes, and multiple gestation. By measurement of maternal serum SP-1 levels alone, 15.2%of the DS pregnancies were detected with a falsepositive rate of 4.3%.This level of detection, which was lower than that observed by Bartels and Lindermann [19881 and by Wald et al. 119891, was comparable to maternal serum hCG alone and MSAFP alone. Combining maternal serum SP-1concentrations with maternal serum AFP, hCG, and maternal age a t the time of delivery, through a logistic multiple regression analysis, it was possible to formulate a risk factor for each patient that was 78.3% predictive of DS with a false-positive rate of 3.4%.Without AFP and hCG included in this model, 3 less pregnancies were detected (Table I). Undoubtedly, inherent biases exist in the present study that may mitigate the predictiveness of this method of combinational analysis. While the model may hold great value for screening programs in the future,
I
the lack of readily availabile assay kits or procedures may limit its present application. Additional studies, including prospective studies, are necessary for adequate assessment of its value to prenatal screening for DS. TABLE I. Number and Proportion of Predicted DS Pregnancies* Down syndrome pregnancies n %
Normal pregnancies n %
AFP only (MOM 2.9) SP-1 only (MOM >2.8) Combined maternal age and SP-1 Combined maternal age, AFP, hCG, and SP-1 *Normal pregnancies n
=
7 5 5 5
4.3 4.3 4.3
7 11 7 33
15.2 23.9 15.2 71.7
4
3.4
36
78.3
6.0
117; DS pregnancies n
=
46.
118
Petrocik et al.
REFERENCES Aldrich JH, Nelson FD (1984): “Linear Probability Logit and Probit Models.” Beverly Hills, California: Sage Publications. Anthony R, Masson GM, Wood PJ (1980): The radioimmunoassay of pregnancy specific B1-glycoprotein in plasma as a pregnancy test for subfertile women. Br J Obstet Gynaecol 87:496-500. Bartels I, Lindemann A (1988): Maternal levels of pregnancy-specific B1-glycoprotein (SP-1) are elevated in pregnancies affected by Down’s syndrome. Hum Genet 80:46-48. Bartels I, Thick M, Bogart MH (1990): Maternal serum hCG and SP1in pregnancies with fetal aneuploidy. Am J Med Genet (in press). Bogart MH, Pandian MR, Jones OW (1987): Abnormal maternal serum chorionic gonadotropin levels in pregnancies with fetal chromosome abnormalities. Prenat Diagn 7:623-630. Bogart MH, Golbus MS, Sorg ND, Jones OW (1989): Human chorionic gonadotropin levels in pregnancies with aneuploid fetuses. Prenat Diagn 9:379-384. Canick JA, Knight GJ, Palomaki GE, Haddow JE, Cuckle HS, Wald NJ (1988):Low second trimester maternal serum unconjugated oestriol in pregnancies with Down’s syndrome. Br J Obstet Gynaecol 95:330-333. Cuckle HS, Wald NJ, Lindebaum RH (1984): Maternal serum alpha fetoprotein measurement: A screening test for Down syndrome. Lancet 1:926-929. Dati F, Grenner G, Luben G, Kapmeyer W, Sieber A, Bohn H, Bellmann 0 (1982):Comparison of enzyme-immunoassays for SP1 and AFP with other immunological methods. La Ricerca Clin Lab. Doran TA, Cadesky K, Wong PY, Mastrogiacomo C, Capello T (1986): Maternal serum alpha fetoprotein and fetal autosomal trisomies. Am J Obstet Gynecol 154:277-281. Eiermann W, Albrich W, Dati F, Leis D, Eicher W (1981): Enzymimmunoassay: Klinische Anwendung. I. Schwangerschaftsfruhdiagnose. Geburtsh Frauenheilkd 41:404-406.
Goodwin BA, Heuther CA (1987):Revised estimates and projections of the Down syndrome births in the United States, and the effects of prenatal diagnosis utilization, 1970-2002. Prenat Diagn 7:261-271. Grudzinskas G, Gordon YB, Jeffrey D, Chard T (1977): Specific and sensitive determination of pregnancy-specific B1-glycoprotein by radioimmunoassay. Lancet 1:331-335. Horne CHW, Towler CM, Pugh-Humpreys RGP (1976):Pregnancy specific B1-glycoprotein-A product of the syncytiotrophoblast. Experientia 32:1197-1201. Mendenhall W (1979): “Introduction to Probability and Statistics.” 5th Ed. North Scituate, Massachusetts: Duxbury Press. Merkatz IR, Nitowsky HM, Marci JN, Johnson WE (1984): An association between low maternal serum alpha fetoprotein and fetal chromosome abnormalities. Am J Obstet Gynecol 148:886-894. Milunsky A, Wands J , Brambati B, Bonacchi I, Currie K (1988): First trimester maternal serum fetoprotein screening for chromosome defects. Am J Obstet Gynecol 159:209-213. Petrocik E, Wassman ER, Kelly J C (1989): Prenatal screening for Down syndrome with maternal serum human chorionic gonadotropin levels. Am J Obstet Gynecol 161:1168-1173. Wald N, Cuckle H, Densem J (19891: Maternal serum specific beta 1-glycoprotein in pregnancies associated with Down’s syndrome. Lancet 2:450. Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Royston P, Chard T, Haddow J E , Knight GJ,Palomaki GE, Canick JA (1989): Maternal serum screening for Down’s syndrome in early pregnancy. Br Med J 297:883-887. Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Canick JA, Haddow JE, Knight GJ, Palomaki GE (1989): Maternal serum unconjugated oestriol as an antenatal screening test for Down’s syndrome. Br J Obstet Gynaecol 95:334-341.
