OBJECTIVE: To evaluate the performance and use of second-trimester multiple-marker maternal serum screening for trisomy 21 by women who had previously undergone first-trimester combined screening (nuchal translucency, pregnancy-associated plasma protein A, and free β-hCG), with disclosure of risk estimates.
METHODS: In a multicenter, first-trimester screening study sponsored by the National Institute of Child Health and Human Development, multiple-marker maternal serum screening with alpha-fetoprotein, unconjugated estriol, and total hCG was performed in 4,145 (7 with trisomy 21) of 7,392 (9 with trisomy 21) women who were first-trimester screen-negative and 180 (7 with trisomy 21) of 813 (52 with trisomy 21) who were first-trimester screen-positive. Second-trimester risks were calculated using multiples of the median and a standardized risk algorithm with a cutoff risk of 1:270.
RESULTS: Among the first-trimester screen-negative cohort, 6 of 7 (86%) trisomy 21 cases were detected by second-trimester multiple-marker maternal serum screening with a false-positive rate of 8.9%. Among the first-trimester screen-positive cohort, all 7 trisomy 21 cases were also detected in the second trimester, albeit with a 38.7% false-positive rate.
CONCLUSION: Our data demonstrate that a sequential screening program that provides patients with first-trimester results and offers the option for early invasive testing or additional serum screening in the second trimester can detect 98% of trisomy 21–affected pregnancies. However, such an approach will result in 17% of patients being considered at risk and, hence, potentially having an invasive test.
LEVEL OF EVIDENCE: II-2
Second-trimester serum screening, after disclosing results of first-trimester combined screening, detected 98% of trisomy 21 pregnancies, but 17% of patients were screen-positive.
From the *Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center and David Geffen School of Medicine, University of California, Los Angeles, California; †Department of Obstetrics and Gynecology, Wayne State University, Detroit, Michigan; ‡The Biostatistics Center, George Washington University, Washington, DC; §NTD Laboratories, Huntington Station, New York; ¶Baylor College of Medicine, Houston, Texas; ∥Evanston Hospital of Northwestern University Medical School, Evanston, Illinois; **Prentice Women's Hospital of Northwestern University, Chicago, Illinois; ††University of California, Los Angeles–Prenatal Diagnosis Unit, Los Angeles, California; ‡‡Yale University, New Haven, Connecticut; §§Magee-Women's Hospital, Pittsburgh, Pennsylvania; ¶¶British Columbia Women's Hospital, Vancouver, British Columbia, Canada; ∥∥McMaster University Medical Centre, Hamilton, Ontario, Canada; ***Prenatal Diagnosis of Northern California Medical Group, Sacramento, California; and †††Drexel University College of Medicine, Philadelphia, Pennsylvania.
*For other members of the BUN Study Group, see the Appendix.
This study was supported by grants R01 HD31991 and HD32109 from the National Institute of Child Health and Human Development.
Address reprint requests to: Lawrence D. Platt, MD, 6310 San Vicente Boulevard, Suite 520, Los Angeles, CA 90048; e-mail: LPlatt8496@aol.com.
Received February 5, 2004. Received in revised form June 14, 2004. Accepted July 1, 2004.
The options for prenatal screening for trisomy 21 have grown rapidly over the last 20 years from risk assessment based solely on maternal age to the incorporation of multiple markers found in maternal serum and, more recently, ultrasound features of the fetus.1 Screening solely by maternal age yields a detection rate of about 30%, with a 7.5% false-positive rate, whereas screening with maternal serum alpha-fetoprotein (MSAFP), human chorionic gonadotropin (hCG), and unconjugated estriol has an overall detection rate of 63% at a set 5% false-positive rate.2 Others have reported an increased detection rate to about 75% by adding inhibin-A (“quad” test).3
Introduced in the 1990s by Nicolaides et al,4–7 fetal nuchal translucency measurement at 11–14 weeks of gestation, when combined with maternal age, had a trisomy 21 detection rate reported to be as high as 77% (5% false-positive rate). Screening by the 2 maternal serum markers alone, free β-hCG and pregnancy-associated plasma protein A (PAPP-A), in the first trimester detects 60–65% of trisomy 21–affected pregnancies.8 Combining nuchal translucency, free β-hCG, and PAPP-A with maternal age yields approximately an 80–90% detection rate of trisomy 21–affected pregnancies with a 5% false-positive rate, as shown in earlier studies9,10 and more recently by our National Institute of Child Health and Human Development (NICHD)-sponsored, multicenter collaborative study, the First Trimester Maternal Serum Biochemistry and Ultrasound Fetal Nuchal Translucency Screening (BUN) Study.11 This study evaluated the efficacy of this screening model in a large population and assessed the practical concerns of achieving consistent and accurate nuchal translucency measurements in multiple centers among multiple sonographers, which have been voiced by some.12,13 Trisomy 21 detection rate was 85.2%, with a 9.4% false-positive rate, when the risk of an affected pregnancy was 1:270 or higher. Adjusting these figures for a set 5% false-positive rate in a modeled general population, the detection rate decreases to 78.7%. The improved detection rate of first-trimester screening over second-trimester maternal serum screening held true even when taking into account the differences in spontaneous loss rates of trisomic pregnancies between the 2 trimesters.
Although many women having first-trimester screening will choose to act on positive results and have early prenatal diagnostic testing by chorionic villous sampling (CVS), others may opt for additional noninvasive testing in the second trimester. Thus, further assessment regarding trisomy 21 screening is needed to determine 1) the role of second-trimester screening as a continuum after first-trimester screening with disclosure of results to patients and 2) the optimal way to interpret these added results.
Although it is true that some screening models would delay disclosure of first-trimester results until after second-trimester serum testing is completed (integrated testing in which risk estimates are derived by combining maternal age, first-trimester nuchal translucency, and PAPP-A and second-trimester MSAFP, unconjugated estriol, hCG, and inhibin-A likelihood ratios),14 given the available data on the efficacy of first-trimester results alone, physician-directed withholding of information seems to us unfair to the patient, for it precludes earlier decision making. If second-trimester screening adds value, a sequential approach that does not routinely require withholding information may be warranted. Women who have first-trimester screening and choose not to undergo early invasive testing can be offered second-trimester screening as well. However, several uncertainties must be addressed before this approach can be implemented. Specifically, removal of screen-positive trisomy 21 cases in the first trimester will decrease the prevalence of affected pregnancies in the second trimester and, subsequently, the positive predictive value of second-trimester serum screening.15,16 Given that the false-positive rates for both the first- and second-trimester screen must be combined, the resulting rate may be unacceptably high.17 To help address these issues, we describe our experience using data from the NICHD-sponsored BUN Study in women who had both first- and second-trimester screening.
MATERIALS AND METHODS
The screening protocol for this study has been presented in detail elsewhere.11 At 12 participating prenatal diagnostic centers, pregnant women underwent first-trimester screening for trisomies 21 and 18 based on maternal age, PAPP-A, free β-hCG, and nuchal translucency. A trisomy 21 risk of at least 1:270 or a trisomy 18 result of at least 1:150 was considered screen-positive for all first-trimester studies. All screen-positive patients and all women 35 years of age or older were offered an invasive procedure. The protocol additionally required that all patients continuing their pregnancies into the second trimester be offered multiple marker maternal serum screening by MSAFP, total hCG, and unconjugated estriol. The study was approved by the institutional review boards at each of the 12 participating prenatal diagnostic centers.
The distribution parameters of Cuckle18 were used to calculate second-trimester risk in a standardized manner from the multiples of the median (MoM) of the analytes as reported by the individual centers. Maternal age–related a priori risk for trisomy 21 in the second trimester was used for second-trimester multiple-marker maternal serum screening analysis. A risk for trisomy 21 of at least 1:270 was considered screen-positive, although second-trimester serum analyte interpretations reported to women based on local cutoffs differed from this in some cases (eg, in California the state-mandated cutoff is 1:190). First-trimester combined screening results were not incorporated into the second-trimester multiple-marker maternal serum screening risk analysis. All cases were assessed for the presence of aneuploidy by prenatal karyotype if invasive testing was performed, from products of conception in the event of a spontaneous abortion (< 24 weeks of gestation) or by postnatal phenotypic evaluation.
In the primary study 8,216 women with no previous trisomy 21- or trisomy 18–affected pregnancies completed first-trimester screening and had primary outcome available for analysis. Because the purpose of this paper is to describe screening for trisomy 21 exclusively, 11 cases of trisomy 18 have been removed, leaving 8,205 women in the analysis (Fig. 1). There were 813 women (9.9%) who screened positive for trisomy 21 after the BUN test; this group contained 52 of the 61 (85.2%) cases of trisomy 21. Four hundred sixty-four (57.1%) of the BUN screen-positive women had invasive testing and did not complete multiple-marker maternal serum screening. In this group 71.2% (37/52) of the BUN-positive trisomy 21–affected pregnancies were identified. There was an intrauterine fetal demise before 15 weeks in 5 cases, with 1 found to be affected with trisomy 21. Multiple-marker maternal serum screening was not done or results were not available in 164 BUN-positive women who did not have prenatal diagnostic testing. Seven of these patients were later found to have trisomy 21–affected pregnancies. Of the 7,392 women who were screen-negative in the first trimester, 1,134 had CVS or amniocentesis without completing multiple-marker maternal serum screening, and 2 pregnancies were found to be affected with trisomy 21. Multiple-marker screening was not obtained in 15 cases, which were not viable at 15 weeks. Multiple-marker screening was declined or results were unavailable in 2,098 of the remaining BUN-negative cases. None of these pregnancies were affected with trisomy 21. The analysis reported here is based on the 4,325 BUN-screened patients for whom second-trimester multiple-marker screening results from MSAFP, hCG, and unconjugated estriol were available.
We calculated sensitivity and false-positive rates for second-trimester multiple-marker maternal serum screening on the basis of screen-positive and -negative first-trimester screening results and the odds of being affected with a trisomy 21 pregnancy following positive or negative results at each screening. The Spearman correlation coefficient was used to compare first-trimester free β-hCG and second-trimester total hCG results in MoMs. Exact binomial confidence intervals were calculated. SAS 8.0 software (SAS Institute Inc, Cary, NC) was used for analysis.
Table 1 lists the maternal characteristics for 4,325/8,205 (52.7%) BUN study patients who went on to have multiple-marker maternal serum screening after first-trimester combined screening for Down syndrome. Figure 1 gives multiple-marker screening results in 4 bolded boxes in which trisomy 21–affected pregnancies are indicated in parentheses. One hundred eighty (31.1%) of the 579 BUN-positive women not having either CVS or amniocentesis before 15 weeks of gestation had second-trimester multiple-marker screening. One hundred six (58.9%) were screen-negative in the second trimester. None were affected with trisomy 21. Seventy-four women (41.1%) were screen-positive in both trimesters, and this group included 7 cases of trisomy 21. Thus, the odds of having an affected fetus with positive first- and second-trimester screening were about 1:10. Although all 7 trisomy 21 pregnancies screened by multiple-marker maternal serum screening in this group were identified, the false-positive rate was 38.7%.
A total of 4,145 (59.3%) of the 6,994 BUN-negative patients who did not have first-trimester invasive testing had multiple-marker screening reported. Three hundred seventy-four (9.0%) were screen-positive only in the second trimester, among which there were 6 cases of trisomy 21. Therefore, the odds of having an affected fetus in this scenario were 1:61. One trisomy 21 case was screen-negative in both the first and second trimesters, but screen-positive for trisomy 18 in the first trimester. Table 2 lists the marker and risk values for the 9 BUN-negative trisomy 21–affected pregnancies.
For the cohort of patients 35 years of age and older, if only those who were BUN or multiple-marker screen–positive had been offered invasive testing, we estimate that all trisomy 21 cases would have been detected for a procedure rate of 30.6%. The odds of being affected with trisomy 21 with any positive result in this group were 1:19. With the same scenario for women younger than 35 years, 11 of 12 trisomy 21 cases would have been detected (odds of being affected, 1:23). Conversely, of the 1,276 patients 35 years of age and older who were BUN and multiple-marker screen–negative, none had a trisomy 21–affected pregnancy. The odds of being affected with trisomy 21 after negative results in both screening tests in women under 35 were 1:2,494.
Of the 1,276 women, 35 years of age and older, who had negative screening tests in both trimesters, 106 chose to have a second-trimester amniocentesis. None of these pregnancies were affected with trisomy 21 (mean maternal age at delivery 37.6 years, median 37.3, range 35–43). Eighty-eight women aged less than 35 years elected second-trimester amniocentesis despite 2 negative screens, for reasons that included parental anxiety and positive family history. There were no cases of trisomy 21 in these pregnancies. Of the 3 trisomy 21 cases without a positive first- or second-trimester screen for trisomy 21, 2 had a second-trimester amniocentesis for an indication of advanced maternal age without having multiple-marker maternal serum screening drawn before sampling.
Table 3 gives the detection rate and false-positive rate for various multiple marker screening cutoff points for the cohort of women who were BUN-screen–negative and also had multiple-marker screening. There were 7 cases of trisomy 21 in this group, of which 6 were detected (85.7%) with a false-positive rate of 8.9%. Even with a relatively high second-trimester risk cutoff of 1:47, the overall detection rate of sequential analysis would be 94%, with a false-positive rate of about 10%.
This report provides useful information from our NICHD collaborative study11 regarding the potential efficacy of sequential screening following first-trimester screening. With the disclosure of their first-trimester screen results, many women may still request second-trimester multiple-marker maternal serum screening. This approach would provide patients the maximal number of options with which to make informed decisions. Our data demonstrate that such a screening program would improve the detection of trisomy 21–affected pregnancies. Sequential screening had a detection rate of 98%; however, it was accomplished with a screen-positive rate of about 17%.
In the primary study, first-trimester combined screening was presented as a research protocol. However, test results were disclosed with decisions regarding subsequent care left to the discretion of the patient and her physician. This approach limits the analysis of the data reported here. First, in spite of the recommendation for second-trimester multiple-marker screening for continuing pregnancies, subsequent screening was not done or results were unavailable in 47% of BUN-screened patients. We are at a loss to account for this low uptake of the second-trimester screening. Invasive prenatal diagnostic sampling at less than 15 weeks was performed in 7.5% of screening patients. It would appear that access to the first-trimester results may have played a role in the decision making in this small cohort, in that 28% of BUN-screen–positive, compared with 5% of the BUN-screen–negative, patients underwent invasive testing before 15 weeks, forgoing second-trimester multiple-marker screening. Second, trisomy 21 was diagnosed in a proportion of the patients (22/61) in this cohort, and these patients were removed from the population before second-trimester screening could be performed.
Additionally it should be noted that hCG values were used to calculate risks in both trimesters. This will have magnified the effect of this analyte because the correlation coefficient between trimesters in our series for free β-hCG and total hCG was 0.6. In the absence of any statistical adjustment, these limitations affect any definitive conclusions regarding the relative benefits of second-trimester serum screening in this population. Nonetheless, we believe the findings reported do offer pertinent clinical observations that warrant further studies.
Women 35 years of age and older appear to have undertaken the BUN screen for various reasons. Some wanted information to help them decide between CVS and amniocentesis because of perceived differences in risks of the 2 procedures. Despite the fact that these 2 procedures have been demonstrated to be equally safe,19 there remains a preference by some patients to await an amniocentesis if the likelihood of a positive result is low. Others wanted to evaluate their risk more thoroughly before undergoing a procedure in an attempt to avoid having any invasive testing if their screen risk was low. Finally, some women wanted information on the day of their CVS that might alleviate some anxiety while waiting for results, although for these patients, the risk revision they received was based solely on the nuchal translucency information.
Until such time as sufficient data are available in a large cohort evaluating the cost/risk/benefit of sequential or integrated testing, all women having first-trimester screening still need to be made aware that the present standard of care is second-trimester serum screening. Although maximum detection with the lowest false-positive rates may appear to be the driving force for choosing a screening approach, that is not always true. To make screening most effective, it is equally important to understand patients’ preferences and to consider their perceptions of risks based on gestational age and their willingness to use various paradigms, including ultrasonography and other markers that may be identified in the future.
Finally, despite the limitations noted above, these results continue to demonstrate a superior screening efficiency in the use of multiple markers for the detection of trisomy 21 and further support for the need to abandon the use of maternal age alone as a basis for invasive testing.
1. Hobbins JC, Lezotte DC, Persutte WH, DeVore GR, Benacerraf BR, Nyberg DA, et al. An 8-center study to evaluate the utility of mid-term genetic sonograms among high-risk pregnancies. J Ultrasound Med 2003;22:33–8.
2. Cuckle H. Established markers in second trimester maternal serum. Early Hum Dev 1996;47(suppl):27–9.
3. Wald NJ, Huttly WJ, Hackshaw A. Antenatal screening for Down's syndrome with the quadruple test. Lancet 2003;361:835–6.
4. Nicolaides KH, Azar G, Byrne D, Mansur C, Marks K. Fetal nuchal translucency: ultrasound screening for chromosomal defects in first trimester of pregnancy. BMJ 1992;304:867–9.
5. Snijders RJ, Noble P, Sebire N, Souka A, Nicolaides KH. UK multicentre project on assessment of risk of trisomy 21 by maternal age and fetal nuchal translucency thickness at 10–14 weeks of gestation. Lancet 1998;352:343–6.
6. Pandya PP, Brizot ML, Kuhn P, Snijders RJ, Nicolaides KH. First trimester fetal nuchal translucency thickness and risk for trisomies. Obstet Gynecol 1994;84:420–3.
7. Pandya PP, Snijders RJ, Johnson SP, De Lourdes Brizot M, Nicolaides KH. Screening for fetal trisomies by maternal age and fetal nuchal translucency thickness at 10–14 weeks of gestation. Br J Obstet Gynaecol 1995;102:957–62.
8. Krantz DA, Larsen JW, Buchanan PD, Macri JN. First-trimester Down syndrome screening: free beta-human chorionic gonadotropin and pregnancy-associated plasma protein A. Am J Obstet Gynecol 1996;174:612–6.
9. Krantz DA, Hallahan TW, Orlandi F, Buchanan P, Larsen JW Jr, Macri JN. First trimester Down syndrome screening using dried blood biochemistry and nuchal translucency. Obstet Gynecol. 2000;96:207–13.
10. Spencer K, Souter V, Tul N, Snijders R, Nicolaides KH. A screening program for trisomy 21 at 10–14 weeks using fetal nuchal translucency, maternal serum free beta-human chorionic gonadotropin and pregnancy-associated plasma protein-A. Ultrasound Obstet Gynecol 1999;13:231–7.
11. Wapner R, Thom E, Simpson JL, Pergament E, Silver R, Filkins K, et al. First-trimester screening for trisomies 21 and 18. N Engl J Med 2003;349:1405–13.
12. Snijders RJ, Thom EA, Zachary JM, Platt LD, Greene N, Jackson LG, et al. First-trimester trisomy screening: nuchal translucency measurement training and quality assurance to correct and unify technique. Ultrasound Obstet Gynecol 2002;19:353–9.
13. Haddow JE, Palomaki GE, Knight GJ, Williams J, Miller WA, Johnson A. Screening of maternal serum for fetal Down's syndrome in the first trimester. N Engl J Med 1998;338:955–61.
14. Wald NJ, Rodeck C, Hackshaw AK, Walters J, Chitty L, Mackinson AM, et al. First and second trimester antenatal screening for Down's syndrome: the results of the Serum, Urine and Ultrasound Screening Study (SURUSS). Health Technol Assess 2003;7:1–77.
15. Kadir RA, Economides DL. The effect of nuchal translucency measurement on second-trimester biochemical screening for Down's syndrome. Ultrasound Obstet Gynecol 1997;9:244–7.
16. Thilaganathan B, Slack A, Wathen NC. Effect of first-trimester nuchal translucency on second-trimester maternal serum biochemical screening for Down syndrome. Ultrasound Obstet Gynecol 1997;10:261–4.
17. Rozenberg P, Malagrida L, Cuckle H, Durand-Zaleski I, Nisand I, Audibert F, et al. Down’s syndrome screening with nuchal translucency at 12(+0)-14(+0) weeks and maternal serum markers at 14(+1)-17(+0) weeks: a prospective study. Hum Reprod 2002;17:1093–8.
18. Cuckle H. Improved parameters for risk estimation in Down's syndrome screening. Prenat Diagn 1995;15:1057–65.
19. Rhoads GG, Jackson LG, Schlesselman SE, de la Cruz FF, Desnick RJ, Golbus MS, et al. The safety and efficacy of chorionic villus sampling for early prenatal diagnosis of cytogenetics abnormalities. N Engl J Med 1989;320:609–17.
Including the authors, the members of The First Trimester Maternal Serum Biochemistry and Fetal Nuchal Translucency Screening Study are as follows:
Drexel University: M. DiVito and M. McGee; Baylor College of Medicine: A. Burke, J. Dungan; Prentice Women's Hospital of Northwestern: K. DeMarco, R. Sabbagha, and N. Ginsberg; Evanston Hospital of Northwestern: P Chilis and K. Blum; Cedar-Sinai Medical Center and David Geffen School of Medicine at UCLA: D. Carlson (deceased), J. Williams, D. Krakow, C. Walla, R. Falk, and K. Wendt; Magee-Women's Hospital: L. M. Hill and K. A. Ventura; University of California, Los Angeles–Prenatal Diagnosis Unit: S. Beverly; Wayne State University: P. Devers; Yale University: S. Turk; McMaster University Medical Center: M. L. Beecroft; British Columbia Women's Hospital: S. Soanes; Prenatal Diagnosis of Northern California Medical Group; K. Kahl; and NTD Laboratories, Huntington Station: T. Hallahan. Cited Here...