Introduction

Advanced maternal age (AMA), usually defined as ≥35 years of age at delivery, is an important risk factor for fetuses with chromosomal abnormalities. Since the 1970s when genetic amniocentesis became widely available in clinical practice, amniocentesis has been offered routinely to all pregnant women with AMA [1]. Because this strategy has a high false-positive rate, the increasing use of genetic amniocentesis with the increase in the proportion of pregnancies complicated by AMA is impacting problems, such as unnecessary invasive procedures and their associated fetal loss [2, 3]. Moreover, with the introduction of maternal serum biomarkers and sonographic screening for Down syndrome, an increasing number of pregnant women with AMA prefer to have such a screening test as a risk modifier to potentially avoid amniocentesis [4]. The American College of Obstetrics and Gynecology has recently published a Practice Bulletin stating that “maternal age of 35 years alone should no longer be used as a cutoff in determining who should be offered invasive testing.” [5].

Despite such a guideline, however, in practice direct amniocentesis still appears to be offered frequently to pregnant women with AMA. According to a questionnaire survey of 968 ACOG fellows by Driscoll et al. [6] in 2009, the majority of obstetricians continue to use AMA as a cutoff for offering a diagnostic test. Accordingly, for more accurate genetic counseling, it is necessary to accumulate experiences and data related to various screening tests for women with AMA.

Screening that combines first-and second-trimester screening, such as integration or sequential testing, has been reported as the most improved method of screening [7], and recent changes in the guidelines of genetic counseling for pregnant women with AMA appears based on the high detection rate of these screening tools. In practice, however, the proportion of quadruple screening is high among screening tests because several countries, including Korea, limit insurance coverage for the cost of screening tests to second trimester screening. The quadruple screening test is a preferred option for pregnant woman who present for the first time in the second trimester of pregnancy [8]. Therefore, for pregnant women with AMA, a study is needed to determine if a quadruple screening test can also be an effective option to avoid direct amniocentesis.

Thus, this study determined the screening performance of the quadruple test in pregnant women ≥35 of age in comparison with pregnant women <35 years of age. Specifically, we studied pregnant women ≥40 as a separate group to provide more individualized genetic counseling to pregnant women with AMA, the number of whom is increasing steadily.

Materials and methods

We investigated the medical records of all Korean pregnant women who had received a quadruple screening test for Down syndrome screening at 15–21 weeks gestation at 1 of 4 hospitals (Seoul St. Mary Hospital, Yeouido St. Mary Hospital, St. Paul Hospital, and Cheongwha Women’s Medical Center) between January 2005 and December 2009. Cases in which the fetal chromosomal outcomes were unknown due to loss of follow-up during the peripartum period (n = 1,013) and 4 cases of fetal demise in utero that did not undergo fetal chromosomal analysis were excluded; thus, 9,435 cases were selected as the subjects for this study. All of the pregnant women were given information on the several types of screening tests, and pregnant women with AMA were also given explanations about the options of direct fetal chromosomal analysis. The subjects of this study were limited to pregnant women who elected quadruple screening tests.

Gestational age was estimated by the last menstruation if the cycle had been regular, or by ultrasonographic scan. Maternal age was calculated for the expected delivery date. Maternal serum levels of α-fetoprotein, total human chorionic gonadotrophin, unconjugated estriol, and inhibin A were determined using the Unicel™ Dxl 800 Access Immunoassay System with reagents (Beckman Coulter®, Inc., Fullerton, CA, USA). All data were analyzed by the HIT program (Hamchoon Inc., Seoul, Korea). Approval was obtained from the Institutional Review Board of the Catholic University of Korea, College of Medicine.

Data were obtained from the medical records regarding the expected risk of Down and Edward syndromes, and the expected maternal age at the time of delivery. To determine the presence of fetal chromosomal abnormalities, we examined the results of fetal chromosomal analysis in pregnant women who had amniocentesis due to positive screening tests. For those pregnant women who did not undergo prenatal fetal chromosomal analysis, we examined the neonatal medical records 1 month after birth. According to maternal age, the subjects of this study were divided into three groups, as follows: <35 years of age; 35–39 years of age; and ≥40 years of age. The screening performance of the quadruple test was compared among the three groups. A positive screen was defined as a case in which the expected risk of Down syndrome was ≥1:270 or the expected risk of Edward syndrome was ≥1:100, and the screening performances of Down and Edward syndromes, as well as all aneuploides, were calculated. For evaluating diagnostic performance, we constructed ROC curves and calculated the area under the curve for each of the three groups. The optimized screening cut-off value was selected using the Youden index. A p value <0.05 was considered statistically significant. The analysis software used was the Statistical Package for Social Science (version 12.0; SPSS, Inc., Chicago, IL, USA).

Results

Of 9,435 pregnant women who underwent a quadruple screening test, the mean age at delivery was 32.4 ± 3.8 years, and the mean gestational age when the quadruple screening test was performed was 16.3 ± 1.1 weeks. According to the age at delivery, 6,922 of the pregnant women were <35 years of age, 2,284 were 35–39 years of age, and 229 were ≥40 of age (Fig. 1). The number of pregnant women who had positive quadruple screening tests was 911 (9.7%); 460 (6.6%) had positive quadruple screening tests in the group <35 years of age, 367 (16.1%) had positive quadruple screening tests in the group 35–39 years of age, and 84 (36.7%) had positive quadruple screening tests in the group ≥40 years of age. Among the 911 pregnant women, 568 (62.3%) had genetic amniocenteses, including 304 (66.0%) in the group <35 years of age, 218 (59.4%) in the group 35–39 years of age, and 46 (54.8%) in the group ≥40 years of age. Aneuploides were detected in 24 cases (2.5 per 1,000 pregnancies), 14 of which (1.6 per 1,000 pregnancies) were Down syndrome. With an increase in maternal age, the incidence of aneuploides is also increased. Thus, the incidence of aneuploides was highest (2.6%) in the group ≥40 years of age. Table 1 shows the results of the screening test in each age group. There was 1 case of Down syndrome and 1 case of Turner syndrome, which were negative in the screening test, but shown to be false-negative results. The case with Turner syndrome had a 5-mm nuchal thickness, thus prenatal diagnosis through amniocentesis was possible.

Fig. 1
figure 1

Distribution of pregnant women by maternal age

Full size image
Table 1 Outcome of quadruple screening test according to 3 groups of maternal age
Full size table

Table 2 shows the performance of the quadruple screening test by age group. When the risk cut-off value of Down syndrome was set to 1:270 and the risk cut-off value of Edward syndrome was set to 1:100, the detection rate for Down syndrome or Edward syndrome was 80% in the pregnant women <35 years of age, but 100% in those ≥35 of age. The screening performance was significantly different according to age. The false-positive rate was 6.6% in the group <35 years of age, 15.8% in the group 35–39 years of age, and 35.3% in the group ≥40. Including 6 cases of other aneuploides detected through screening tests, the detection rate of quadruple screening for all aneuploides was 91.6%.

Table 2 Rates of detection and false-positive results for screening of aneuploidies including Down and Edward syndrome on quadruple screening test
Full size table

In the attempt to identify the optimal risk cut-off value to detect Down syndrome in the second trimester of pregnancy, the receiver operating characteristic (ROC) curve analysis of the quadruple screening test was performed for the group 35–39 years of age and the group ≥40 (Fig. 2). The area under the receiver operating characteristic curve (AUROC) was 0.946 (95% confidence interval [CI], 0.908–0.984, p < 0.001) in the group 35–39 years of age, and 0.976 (95% CI, 0.954–0.998, p < 0.001) in the group ≥40 of age. The optimal cut-off with the highest sensitivity and specificity was 1:160 (sensitivity, 83.3%; false-positive rate, 8.8%) in the group 35–39 years of age, and 1: 20 (sensitivity, 100%; false-positive rate, 5.9%) in the group ≥40 of age.

Fig. 2
figure 2

Receiver operating characteristic (ROC) curve of quadruple screening for Down syndrome in women 35–39 years of age (left) and women ≥40 years of age (right). In the group aged 35–39 years, area under the ROC (AUROC) = 0.946, 95% CI: 0.908–0.984, p < 0.001. In the group aged 40 or older, AUROC = 0.976, 95% CI: 0.954–0.998, p < 0.001

Full size image

Discussion

This study showed that the quadruple test is a satisfactory serum screening method before making a decision about direct genetic amniocentesis in pregnancies complicated by AMA. In this study the detection rate of quadruple screening for Down or Edward syndrome was 80% in women <35 years of age and 100% in women >35 years of age. Also, for all aneuploides, the detection rate of Down and Edward syndromes was 88.9 and 100%, respectively. Because the incidences of chromosomal abnormalities are very low, and the sample size of pregnancies complicated by AMA was relatively small in this study, it may be prudent to consider that the actual detection rate in clinical practice may be <100%. Nevertheless, this study highlights that the quadruple screening test is an efficient screening method for Down syndrome and women of AMA in comparison with women <35 years of age.

We found that not only the detection rate, but also the false-positive rate was different according to maternal age. That is, the false-positive rate was 6.6% in women <35 of age, but increased significantly to 15.8% in women 35–39 years of age, and 35.3% in women 40–45 years of age. Such an impact of maternal age on the detection and false-positive rates of screening tests for Down syndrome has been reported in a few publications [9, 10]. Using the triple marker approach, Reynolds et al. [9] reported that the detection and false-positive rates were 56 and 6.1%, respectively, at 30 years of age, but 91.6 and 40.9%, respectively, at 40. Moreover, Spencer et al. [10] pointed out that the detection rate of combined tests in the first trimester was 84% with a false-positive rate of 4% at 30 years of age, and 100% with a false-positive rate of 67% at 49 years of age. Despite these reports, however, there are few health care professionals in charge of antenatal care who are aware of differences in the detection and false-positive rates according to maternal age and consider the differences in counseling pregnant women [11, 12]. The results of this study confirmed that the quadruple screening test also has detection and false-positive rates in women of AMA as high as other types of screening tests for Down syndrome. The quadruple screening test may provide more accurate genetic counseling for patients who have false-positive rates as high as 35%; in the case of 40-year-old pregnant women, the risk of Down syndrome is very low if the result of quadruple screening is negative.

Given such a high false-positive rate in the AMA group, some argue that the risk cut-off value for the AMA should be reset to correct the high false-positive rate. If the optimal cut-off value of the quadruple screening test was set by age for the subjects of this study, it is proper to use 1:160 (sensitivity, 83.3%; false-positive rate, 8.8%) for those 35–39 years of age and 1:20 (sensitivity, 100%; false-positive rate, 5.9%) for those ≥40 years of age. However, considering the fear of a false-negative result, the considerable detection rate of other aneuploides through screening tests, and the practice of offering amniocentesis to all AMA women for >3 decades, lowering the risk cut-off value for pregnant women of AMA requires further discussions and consensus.

Despite changes in the guidelines for screening for Down syndrome in pregnant women of AMA, many pregnant women of AMA choose direct amniocentesis and a large number of health care professionals recommend direct amniocentesis [6, 13]. In addition to concerns over the accuracy of screening tests, another major reason for the choice is anxiety about the risk of chromosomal abnormalities other than Down syndrome. Maternal age-dependent chromosomal abnormalities include trisomies 21, 18, and 13, XXX, and XXY. XYY, 45,X, triploid karyotypes, and structural abnormalities are known to be unrelated to maternal age [14–16]. Thus, chromosomal abnormalities that women of AMA should be concerned about are XXX, XXY, trisomy 13, Down syndrome, and Edward syndrome, but the incidence is quite low (0.54, 1.04, and 1.57 per 10,000 live births) [17, 18]. In addition, it is known that 50% of other chromosomal abnormalities can be detected by a serum screening test, and such information needs to be provided to pregnant women of AMA in genetic counseling. In the current study, among 2,513 pregnant women of AMA, there were only 2 cases of Turner syndrome and 1 case of 47, XXX. No cases, such as trisomy 13 and XXY, were detected. The 47,XXX case and one of the cases of 45,X were diagnosed through screening tests, and the other case of Turner syndrome had an abnormal sonographic finding that became an indication for prenatal chromosomal analysis. In many cases, however, 47,XXY and 47,XXX were not detected immediately after birth, so we should consider the possibility of undetected cases.

Recently, the proportion of pregnancies in mother with AMA is increasing very rapidly. In Korea, the proportion of pregnant women of AMA has also increased steadily (from 6.2% of all pregnant women in 1999 to 15.4% of all pregnant women in 2009). Therefore, if all pregnant women of AMA and pregnant women <35 of age who have a positive serum screening test (approximately 5% of pregnant women <35 undergo amniocentesis), approximately 1 of every 5 pregnant women will undergo amniocentesis in Korea. Considering the cost of amniocentesis and amniocentesis-related risk of fetal loss, screening strategies based only on maternal age should no longer be used.

Recently, screening that combines first- and second-trimester screening is known to be the most accurate screening test for Down syndrome. Thus, it is most desirable for pregnant women of AMA to choose the combined test among screening tests. In this situation, the reason why the authors focused on the quadruple screening test was to provide data to determine whether or not the quadruple screening test can be an adequate choice for pregnant women of AMA. The results of this study suggest that under conditions in which first trimester screening test is not available, a quadruple screening test is a better choice than direct amniocentesis for pregnancies in mother with AMA. Genetic counseling regarding detection and false-positive rates of serum screening tests need to be individualized according to maternal age, especially in pregnant women of AMA.