The impact of the maternal serum screening program on Down syndrome birth trend in Taiwan

• Main Authors: Yih-Shing Kuo, 郭義興
• Other Authors: Fon-Jou Hsieh
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/82822323174737706853

碩士 === 國立臺灣大學 === 臨床醫學研究所 === 91 === Background Down’s syndrome (DS) is the most common chromosomal abnormality in Taiwan with an overall birth prevalence of 1 in 848 births. The birth of an infant with DS causes great emotional and economic burden for affected families. Furthermore, the lifetime cost per case of DS up to 35 years old was estimated to be 3,780,000 to 4,470,000 NT dollars. Since 1985, we have offered amniocentesis for pregnant women with advanced maternal age (aged 35 and over at the delivery). However, only 25-30 per cent of all cases of DS could be identified in this old age population. The remaining 70-75 per cent of Down cases occur in pregnant women younger than 35. Maternal serum biochemical screening was initiated from the observation of an association between low maternal serum alpha-fetoprotein (AFP) and DS pregnancies in either whites and Asians. DS pregnancies were subsequently found associated with low level of unconjugated estriol (uE3). High maternal serum human chorionic gonadotropin (hCG) and its b-subunits in either Caucasians or Asians were also found to be markers of DS. Second trimester maternal serum screening was then developed and widely adopted by combining these biochemical markers and maternal age. Serum screening, being much less invasive, can be performed for a greater number of pregnant women of all ages and has the estimated potential of detecting up to 60 percent of pregnancies with DS. The maternal serum screening for DS during the second trimester was introduced into Taiwan since 1994 and became widely applicable thereafter. Because the maternal serum screening is not covered by health insurance of the government, the practice is unrestricted and became rapidly and widely accepted as a part of routine antenatal care. However, there is no available data to assess the impact of maternal serum screening on DS birth trend during recent years in Taiwan. The purpose of this study was to determine whether the use of large-scale serum screening in Taiwan had a measurable impact on the epidemiology of DS by comparing DS births in the period before and after the implementation of the test. Materials and Methods The data from the “National Birth Defects Registration and Notification system” of Department of Health in Taiwan from 1993 to 2001 were retrieved. A total of 1,331,616 deliveries were collected including 840 cases of DS with karyotyping confirmation. Cases of DS included live-births and stillbirths (spontaneous abortions or terminations of pregnancy after prenatal diagnosis) without regarding whatever the gestational age. Multiple pregnancies were not excluded. The crude birth rate (per 10,000 deliveries), crude live-birth rate (per 10,000 live-births) and the ratio of live-birth and stillbirth to total DS were analyzed year by year for understanding the change of DS birth trend in recent 9 years in Taiwan. Those of isolated cleft palate were also analyzed as internal control variable. The prenatal diagnosis for isolated cleft palate was mainly by ultrasound screening and usually difficult. Even diagnosed prenatally, subsequent termination was usually not done in this condition. There was no reason to expect any change in the birth trend of isolated cleft palate during the study period. Since the prenatal diagnosis policy for DS is age-dependent, we analyzed these data separately by two maternal age classes at the time of delivery (younger than 35 years vs. 35 years or more). Amniocentesis was initially offered for women 35 years or more. Serum biochemical screening was provided for those aged younger than 35 years of age. In Taiwan, most of maternal serum screening for DS during second trimester adopted the double test, including AFP with total b-hCG or AFP with free b-hCG. The database of “Demographic Fact Book” from Ministry of the Interior, Taiwan from 1991 to 2001 were retrieved and analyzed for understanding the age structure of pregnant women in Taiwan. The database of “Amniocentesis in Pregnant Mothers Aged 34 and Above” from the Bureau of Health Promotion, Department of Health, Taiwan from 1987 to 2001 were also analyzed for understanding the amniocentesis acceptance rates in elderly mothers. Independent binominal analysis for difference between population proportions was used for testing the significance of differences between the percentages of live-births and stillbirths to total DS before and after the implementation of the serum screening test. The effects of the change in maternal age and prenatal diagnosis on the epidemiology of DS during the study period are also analyzed. Results The registration rate of our “National Birth Defect Registration and Notification Systems” increased progressively from 20.3% in 1993 to 96.9% in 2001. In the last 4 years of our study period, the registered births represented over 70% of total births. With increased registration rate, the registered numbers of DS increased progressively. There were 808 registered cases of DS with maternal age data available, including 37.4% (302/808) of cases born by mothers with age greater than or equal to 35 years old and 62.6% (506/808) by those with age younger than 35 years old, and 79.5 per cent of live-births with DS were born to mothers younger than 35 years. This points out the importance of screening in the younger age class. Among the general population in Taiwan, elderly pregnant women aged over 35 represent 4.79% of all mothers in 1992 and it increased steadily to 8.93% in 2001. The amniocentesis acceptance rate in elderly mothers aged over 34 increased progressively from 1987 to 2001. During 1993 and 2000, a three-fold increase (25.3% to 75.8%) was noticed. In spite of an upward shift in the maternal age distribution, the average crude birth rate of DS in Taiwan remained around 6.58 per 10,000 births with slight decrease over the last four years of our study period. In the mean time, there was a marked decrease in the crude live-birth rate of DS during 1995-2001 compared with that of 1993-1994. The crude live-birth rate of DS was 5.45 per 10,000 live-births in 1993-1994. It decreased progressively in 1995-1996 then reached an average of 1.75 per 10,000 live-births after 1996 with a 67.9% decrement. In 1993, the number of live-birth DS was greater than that of stillbirth. After the implementation of maternal serum screening for DS in 1994 the number of live-birth DS became less than that of stillbirth after 1995 and the ratio of live-birth to stillbirth DS cases changed from 3.3:1 to 1:2.8 with reversing point occurring around 1995. In 1993, 23.1% of the DS were stillbirth as compared to 67.5% in 2001 (p<0.001). Comparing the percentage of live-birth and stillbirth in DS annually, the percentages of live-birth DS progressively decreased from 76.9% in 1993 to an average of 27.1% after 1996, while that of stillbirth DS progressively increased from 23.1% in 1993 to an average of 72.9% after 1996 with the crossing point also occurring in 1995. Surprisingly, in Taiwan the implementation of maternal serum screening for DS in 1994 was soon followed by a dramatic change in the birth trend of DS. In contrast, the crude live-birth rate of isolated cleft palate did not show a similar decrease during the study period in the same registration system. The percentages of live-birth and stillbirth to total cases of isolated cleft palate did not show the reverse trend as seen in DS but remained steady. The percentages of DS live-births in total DS births decreased progressively from 76.9% in 1993 to an average of 32.2% after 1994 (P<0.001), it maintains nearly a plateau after 1996. Considering maternal age, similar trends were also noted for DS born by mothers aged less and more than 35 years old respectively. The percentage of DS live-births in total DS births in mothers younger than 35 decreased from 78.3% in 1993 to an average of 39.8% after 1994 (p＜0.001), and that for mothers aged 35 or more decreased from 50% in 1993 to an average of 17.4% after 1994 (p = 0.091). The birth weight distribution of the DS stillbirths shows that ninety percent of the DS stillbirths weighed under 1000gm after 1994. About half of the stillbirth DS were born by mothers under 35 in recent years. Discussion The function of a screening test is to concentrate individuals affected by the screened condition into a small group. Typically, maternal serum screening for DS in second trimester concentrates about 60-70% of pregnancies affected by fetal DS into about 5% of the total screened population. The main considerations for providers of serum screening for DS should be minimizing babies with DS missed by the test, and reducing miscarriage due to amniocentesis or chorionic villus sampling. In Taiwan, maternal serum screening for DS was introduced in 1994 and widely accepted thereafter. Screening for DS in the second trimester of pregnancy using biochemical markers has become an established part of obstetric practices. Unlike genetic amniocentesis, serum screening is not covered by insurance. So it is impossible to know exactly how many women undergo the test in Taiwan. Nevertheless, we estimate from various surveys that between 66-85% of pregnancies are screened and the second trimester double test using AFP and total β-hCG or AFP and freeβ-hCG is most commonly offered. The preliminary reports about the effectiveness of serum screening for DS in Taiwan revealed a 59.5% detection rate and 5.9% false positive rate compatible with the results of Caucasian. The average crude birth rates of DS in the region were around 6.58 per 10,000 births with slight decrease over the last four years of our study period, while there was a marked decrease in the crude live-birth rate of DS and the number of live-birth DS after implementation of serum screening in 1994. This is compatible with the reports in literatures that the increase in prenatal diagnosis performed because of abnormal serum markers was associated with an increase over time in the number and the proportion of DS fetuses detected prenatally and a decrease in the prevalence of DS among live-birth. Although the percentage of elderly mothers aged over 35 increased progressively, most pregnancies occurred around age 27-28 over the audit period. Age factor may not be the main cause for the Down’s birth trend found in our study. Our registration form did not include the modalities of prenatal diagnosis. Whether the stillbirth DS cases were terminated artificially was also not known. We estimate with confidence that almost all the DS stillbirths in our series are diagnosed prenatally as revealed by their birth weights. Ninety-five percent of DS stillbirths in our series weighed under 1500gm. Actually most of them (90 %) were under 1000gm. A large national registry of genetic amniocentesis including 108,600 cases in Taiwan reported that advanced maternal age was the most common indication (56.6%) and 26.6% are indicated by positive maternal serum screenings. Also the amniocentesis acceptance rate of elderly mothers reached 75.6% in 2000. Of course, the high amniocentesis acceptance rate in elderly mothers is one of the major reasons in the reduction of DS live-births observed in our series. As seen in Figure 10, almost half of the DS stillbirths were born by mothers younger than 35 in recent years. As far as we know, there were no major changes in the methods of prenatal diagnosis of DS in recent years except the introduction of maternal serum screening. Thus we may conclude that half of the prenatally diagnosed DS, i.e., the DS stillbirths born by mothers younger than 35, can be attributed to maternal serum screening. Also, as seen from the change of DS live-birth percentage in younger mothers, it decreased from 78.3% in 1993 to an average of 39.8% after 1994. This 49.2% reduction of DS live-birth percentage in younger mothers should be attributed to maternal serum screening. Moreover, in Taiwan many elderly mothers hesitated to take amniocentesis directly, they received serum screening first instead and undertook amniocentesis after a positive result. Thus, some of the stillbirth DS in elderly mothers can also be attributed to the maternal serum screenings. However, the live-birth percentage and the prevalence of live babies with DS remained unchanged with even a small increase after 1997. The possible causes may be the followings: (1) the limitation of the serum screening test itself, (2) refusal of screening test by some younger pregnant women, (3) lack of proper prenatal care in some pregnant population such as immigrants from China and South Asia. In summary, the policy of prenatal diagnosis program including amniocentesis for pregnant women aged 35 or more and the wide spread application of maternal serum screening for DS in younger women was responsible for the marked decrease in the live-births of DS in Taiwan and about 50% of this decrement can be attributed to the implementation of maternal serum screening. Key words: Down’s syndrome/human chorionic gonadotropin/live-birth/maternal age/maternal serum screening