OBJECTIVE: To evaluate the performance of first- and second-trimester screening methods for the detection of aneuploidies other than Down syndrome.
METHODS: Patients with singleton pregnancies at 10 weeks 3 days through 13 weeks 6 days of gestation were recruited at 15 U.S. centers. All patients had a first-trimester nuchal translucency scan, and those without cystic hygroma had a combined test (nuchal translucency, pregnancy-associated plasma protein A, and free β-hCG) and returned at 15–18 weeks for a second-trimester quadruple screen (serum alpha-fetoprotein, total hCG, unconjugated estriol, and inhibin-A). Risk cutoff levels of 1:300 for Down syndrome and 1:100 for trisomy 18 were selected.
RESULTS: Thirty-six thousand one hundred seventy-one patients completed first-trimester screening, and 35,236 completed second-trimester screening. There were 77 cases of non–Down syndrome aneuploidies identified in this population; 41 were positive for a cystic hygroma in the first trimester, and a further 36 had a combined test, of whom 29 proceeded to quadruple screening. First-trimester screening, by cystic hygroma determination or combined screening had a 78% detection rate for all non–Down syndrome aneuploidies, with an overall false-positive rate of 6.0%. Sixty-nine percent of non–Down syndrome aneuploidies were identified as screen-positive by the second-trimester quadruple screen, at a false-positive rate of 8.9%. In the combined test, the use of trisomy 18 risks did not detect any additional non–Down syndrome aneuploidies compared with the Down syndrome risk alone. In second-trimester quadruple screening, a trisomy 18–specific algorithm detected an additional 41% non–Down syndrome aneuploidies not detected using the Down syndrome algorithm.
CONCLUSION: First-trimester Down syndrome screening protocols can detect the majority of cases of non-Down aneuploidies. Addition of a trisomy 18–specific risk algorithm in the second trimester achieves high detection rates for aneuploidies other than Down syndrome.
LEVEL OF EVIDENCE: II
Both first- and second-trimester screening protocols are effective for detection of trisomy 18, and to a lesser degree, trisomy 13, triploidy, and Turner syndrome.
From the 1Royal College of Surgeons in Ireland; 2Columbia University, New York, New York; 3Brown University, Providence, Rhode Island; 4University of Utah, Salt Lake City, Utah; 5Swedish Medical Center, Seattle, Washington; 6William Beaumont Hospital, Royal Oak, Michigan; 7University of Texas Medical Branch at Galveston, Galveston, Texas; 8Mount Sinai Medical Center, New York, New York; 9Montefiore Medical Center, Bronx, New York; 10University of Colorado Health Sciences Center, Denver, Colorado; 11Tufts University, Boston, Massachusetts; 12New York University Medical Center, New York, New York; 13University of North Carolina, Chapel Hill, North Carolina; and 14DM-STAT, Boston, Massachusetts.
Supported by grants from the National Institutes of Health and the National Institute of Child Health and Human Development (RO1 HD 38652 and M01 RR 00054).
Corresponding author: Dr. Fionnuala M. Breathnach, MRCOG, Department of Obstetrics and Gynaecology, Royal College of Surgeons in Ireland, Rotunda Hospital, Parnell Square, Dublin 2, Ireland; e-mail: firstname.lastname@example.org.
Financial Disclosure Dr. Lambert-Messerlian is a consultant for Beckman-Coulter (Fullerton, CA). Dr. Cuckle holds a directorship with Genome, Ltd. (Leeds, UK) and Logical Medical Systems, Ltd. (London, UK). He also received a paid consultancy from Perkin-Elmer Life Sciences (Wallac Oy, Finland). The other authors have no potential conflicts to disclose.
Second-trimester screening using maternal age and the quadruple serum markers alpha-fetoprotein (AFP), total hCG, unconjugated estriol (E3), and inhibin A has been validated as an effective screening tool for Down syndrome.1–4 First-trimester screening, using nuchal translucency combined with maternal age and the serum markers pregnancy-associated plasma protein A (PAPP-A) and free β-hCG has been demonstrated in several large studies to have comparable or greater test performance.5,6 Fetal aneuploidy associated with positive prenatal screening is not limited, however, to Down syndrome. Indeed, among chromosomally abnormal fetuses with increased nuchal translucency, approximately half are affected by conditions other than Down syndrome.7
Both first- and second-trimester screening programs have been expanded to include risk assignment for trisomy 18.8–10 Targeted screening algorithms have not been developed for chromosomal abnormalities other than Down syndrome and trisomy 18, although it has been suggested that a trisomy 13 risk might be calculated.11 The construction of such algorithms would require recognition of a characteristic pattern for each condition using the appropriate combination of markers. On a background of very limited description of such patterns for trisomy 13, Turner syndrome, and triploidy, the extent to which these conditions can be detected using existing Down syndrome screening algorithms has not been established.
The First and Second Trimester Evaluation of Risk (FASTER) trial,2 a prospective multicenter study, was designed to compare a range of serum screening protocols, with or without ultrasonographic measurement of nuchal translucency, for fetal aneuploidy at different gestational ages. The objective of the current study was to evaluate the effectiveness of these protocols for the detection of aneuploidy other than Down syndrome.
MATERIALS AND METHODS
Fifteen U.S. centers participated in this study from October 1999 to December 2002. Written informed consent was obtained from participants in conjunction with institutional review board approval at each recruiting center. The inclusion criteria were a maternal age of 16 years or older, pregnancy with a singleton live fetus, and a fetal crown-rump length of 36 mm to 79 mm (consistent with a gestational age of 10 weeks 3 days through 13 weeks 6 days at study entry).12 A diagnosis of multiple gestation or of anencephaly led to exclusion from the study. All patients had a first-trimester nuchal translucency scan and those without cystic hygroma had a combined test (nuchal translucency, PAPP-A, and free β-hCG, with maternal age). Wherever septated cystic hygroma was diagnosed, this automatically constituted a positive first-trimester screen, and serum samples were not obtained.13
Study participants returned at 15–18 weeks of gestation for second-trimester screening. At that time, an independent second-trimester risk was calculated from measurements of serum AFP, hCG, unconjugated E3, and inhibin A, together with maternal age. For calculation of second-trimester trisomy 18 risk, the former three serum analytes were used.
Ultrasonography to assess nuchal translucency was performed according to a standardized protocol by specially trained ultrasonographers.14 Full details regarding the ultrasound and quality assurance protocols have been published previously.2 Measurements of markers were converted into multiples of the median (MoM) for gestational age, adjusted for maternal weight, type 1 diabetes, and race. Nuchal translucency median data were center-specific, and the mean of three measurements was used for calculation of MoM values. The risk for aneuploidy was estimated by multiplying the maternal age-specific odds of an affected infant15 by the likelihood ratio for each marker obtained from the overlapping Gaussian distributions of affected and unaffected pregnancies.
For Down syndrome, a positive result from first-trimester or second-trimester screening was defined as a risk at the end of pregnancy (40 weeks) of 1 in 150 or 1 in 300, respectively. For trisomy 18, the cutoff level was 1 in 100. Results were not provided to patients until after all screening tests were complete and patients with positive results from either first-trimester or second-trimester screening were offered formal genetic counseling and the option of amniocentesis for genetic analysis.
The higher first-trimester cutoff was chosen for the original trial as a patient-safety step used to ensure that the overall population screen-positive rate would not exceed 8%. For the purpose of the current analysis, a Down syndrome cutoff of 1 in 300 was selected in both trimesters because this cutoff level is closer to what is used in clinical practice. Trisomy 18 risk was also calculated using first-trimester markers and a risk cutoff of 1 in 100. The distribution parameters have been previously described16 and are included in the Appendix.
The following screening tests were evaluated: measurement of first-trimester nuchal translucency alone; first-trimester combined screening (nuchal translucency plus PAPP-A and free β-hCG) using both Down syndrome and trisomy 18 algorithms, second-trimester quadruple screening for Down syndrome, and second-trimester triple screening (AFP, hCG, and unconjugated E3) for trisomy 18. For all tests, the calculated risk took into account maternal age. For each test the observed detection and false-positive rates were computed together with the 95% confidence interval (CI).
As a means of maximizing case ascertainment, copies of fetal and pediatric medical records were submitted for review by a single pediatric geneticist in all cases in which a possible fetal or neonatal medical problem was suspected, in all cases with a positive screening test result but without karyotype results, and in a 10% random sample of all other cases. Fetal chromosome status was determined by amniocentesis, by sampling neonatal cord blood in cases with a positive screening test result in which the mother declined amniocentesis, or by tissue sampling in cases of spontaneous pregnancy loss, pregnancy termination, or stillbirth.
A total of 36,171 patients completed first-trimester screening, and 35,236 completed second-trimester screening. The demographic characteristics of the study groups are outlined in Table 1. There were 77 cases of non–Down syndrome aneuploidy (trisomy 13 or 18, Turner syndrome, or triploidy) in the population of 36,171 women (prevalence 0.2%). This group was composed of the following chromosomal abnormalities: trisomy 18, 28 cases; trisomy 13, 15 cases; Turner syndrome, 26 cases; and eight cases of triploidy. Septated cystic hygroma was detected in 41 (53%) of these pregnancies, and this group was not included in second-trimester analysis, because they did not contribute serum samples. Among nonhygroma cases, spontaneous miscarriage occurred in 7 of 36 (19%) in the interval between first- and second-trimester screening, thus accounting for the fact that 29 of 36 patients contributed second-trimester serum samples. There were no additional aneuploid cases in the 10% review group of randomly selected screen-negative patients. The overall number of aneuploid cases ascertained was similar to the rate expected based on the age distribution of the FASTER population.
Table 2 summarizes the performance of first-trimester nuchal translucency and serum markers in the subgroup of cases in which all markers were obtained (ie, excluding cystic hygroma cases). The number of cases in each subgroup was small so that median and range values are given rather than means and standard deviations, to avoid the undue effect of outliers. The analyte values for these non–Down syndrome aneuploidies that remained viable in the second trimester are also reported. A characteristic pattern of low analyte levels was only observed for trisomy 18 and triploidy.
The comparative data in Table 3 illustrate the efficiency of first-trimester combined screening and second-trimester serum quad screening for detection of non–Down syndrome aneuploidies. Overall, a positive aneuploidy screening result (as defined by either a 1:300 risk using a Down syndrome algorithm or a 1:100 risk using a trisomy 18 algorithm), resulted in a 78% (95% CI 69–87%) detection rate for non–Down syndrome aneuploidies using first-trimester combined and cystic hygroma screening and an overall false-positive rate of 6.0%. Second-trimester quad screening detected 20 of 29 (69%; 95% CI 52–86%) cases through a positive screen for either Down syndrome or trisomy 18. Focusing specifically on a trisomy 18 screening algorithm and using a 1 in 100 risk cutoff, the first-trimester combined test detected 22 of 28 (79%; 95% CI 63–94%) trisomy 18 cases at a 0.3% false-positive rate, whereas second-trimester serum screening detected 13 of 13 (100%) cases at a 0.3% false-positive rate. Focusing specifically on a Down syndrome screening algorithm and using a 1 in 300 risk cutoff, the first-trimester combined test detected 23 of 28 (82%; 95% CI 68–96%) trisomy 18 cases, and the false-positive rate was 6.0%.
Of the eight cases of triploidy, five screened positive with first-trimester combined and cystic hygroma screening. Three of these had septated cystic hygroma and the remaining two miscarried spontaneously before the timing of their second-trimester screen. For both cases of triploidy remaining viable into the second trimester, all analytes were low apart from AFP, and these cases screened positive with second-trimester screening. Second-trimester quadruple screening, using Down syndrome or trisomy 18 risk calculation algorithms, performed less efficiently at screening for trisomy 13 and Turner syndrome.
The addition of a first-trimester trisomy 18 algorithm did not detect any non–Down syndrome aneuploidies that were not detected with first-trimester Down syndrome screening alone. By comparison, with second-trimester quadruple screening, 28% of non–Down syndrome aneuploidies were indentified as screen-positive by the second-trimester Down syndrome risk algorithm, at a false-positive rate of 8.6%, while a trisomy-18–specific algorithm detected an additional 41% (12 of 29) non–Down syndrome aneuploidies, increasing the overall detection rate to 69%, with a 0.3% increment in the false-positive rate.
Among non-hygroma cases, the spontaneous miscarriage rate of 19% (7 of 36) in the interval between first- and second-trimester screening resulted in a disparity between the numbers contributing to the screening performance data in first and second trimesters. Focusing solely on those pregnancies that contributed serum in both trimesters, ie, excluding miscarriage cases, allows direct comparison of identical data sets. This comparison is illustrated in Table 4.
Prenatal screening for Down syndrome is an important component of routine antenatal care. The recognition that second-trimester unconjugated E3, AFP, and hCG levels are low in pregnancies affected by trisomy 188 led to the development of specific algorithms targeted at screening for this condition, and this approach has been incorporated into screening programs for Down syndrome. The 100% detection rate for trisomy 18, at a false-positive rate of 0.3%, using three components of the second-trimester quad screen in our study compares favorably with detection rates of 60–69% reported by others10,17 for this screening strategy. The ability of currently employed Down syndrome and trisomy 18 screening programs to also detect other aneuploid conditions has been investigated.
First-trimester combined screening was also effective for trisomy 18. This provided an 82% detection rate for an overall population false-positive rate of 6.0% when a high-risk result was obtained using a Down syndrome algorithm, a trisomy 18 algorithm, or presence of a cystic hygroma. Because each of these findings would usually result in a chorionic villus sampling being performed, it is reasonable to consider all such cases of trisomy 18 identified in this way as being successfully detected in the first trimester. Focusing purely on the performance of a trisomy 18 screening algorithm in the first trimester (ie, large nuchal translucency, cystic hygroma, low hCG, and low PAPP-A), the detection rate was 79% (22 of 28) at a 0.3% false-positive rate. One limitation of our study is that first-trimester cases identified as having a cystic hygroma were offered chorionic villus sampling immediately and therefore did not contribute serum samples for further marker evaluation.13 Cystic hygroma alone was a powerful predictor of aneuploidy. First-trimester screening without inclusion of cystic hygroma cases led to only 62% detection of aneuploidy at a 5.9% false-positive rate.
The application of a dedicated first-trimester trisomy-18 algorithm to the FASTER data set represents a post hoc analysis. Interestingly, however, as illustrated in Table 4, the addition of this algorithm did not contribute to the detection of aneuploidies other than those identified with trisomy 21 screening. Those cases of trisomy 13, Turner syndrome, and triploidy that screened negative using the first-trimester trisomy 21 algorithm also screened negative where the trisomy 18 algorithm was applied. Similarly, 98% (50 of 51) of those euploid pregnancies that constituted the false-positive group screened positive not only for trisomy 18 but also for trisomy 21 in the first trimester.
When considering the comparative performance of first- and second-trimester screening, the viability bias that results from the detection of aneuploid fetuses that would otherwise have been destined to miscarry should be taken into account as this results in an overstatement of the real detection rate for liveborn aneuploid infants. An additional high lethality rate between the second trimester and term further distorts the detection rates that are based on second-trimester cases. Therefore, the detection rates of viable-to-term cases may be much smaller than our estimates.
Across first and second trimesters, detection rates among the trisomy 13 cases using serum markers were not as impressive as for trisomy 18. Among the 15 cases of trisomy 13 in this study, six were identified with cystic hygroma. Of the remaining nine cases, only four were identified using nuchal translucency, PAPP-A, and free β-hCG in the first trimester, and only three of the seven cases that were still viable in the second trimester were identified using quad serum markers. The finding of low levels of PAPP-A and free β-hCG in the first trimester (median MoM values 0.36 and 0.41, respectively), a pattern shared with trisomy 18, is consistent with another studies.11 The similar analyte patterns that trisomies 13 and 18 share in the first trimester essentially would argue against the construction of a separate first-trimester algorithm for each, a fact recognized by Spencer et al,11 who have rather devised a merged algorithm for trisomies 13 and 18. Using a combined first-trimester screening approach, this group reported a 95% detection rate for these chromosomal defects at 11–14 weeks of gestation, with a false-positive rate of 0.3% where a risk cutoff of 1 in 150 was used.
The fact that the majority of Turner syndrome cases in our study (73%) had septated cystic hygroma contributed to the high detection rates in the first trimester for this condition. The observed second-trimester serum analyte pattern for Turner syndrome described by others18,19 comprises low or normal concentrations of AFP, low unconjugated E3, high hCG, and high inhibin A, a pattern shared with Down syndrome. This pattern, however, is typical of Turner syndrome with hydrops. An alternative analyte pattern, with low marker levels reflective of a trisomy 18 pattern, is reported to be more typical of Turner syndrome without hydrops.20 Obtaining a positive screen in the nonhydropic cases is reliant upon serum analytes, and our data show that second-trimester quadruple Down syndrome screening performed poorly for detecting these cases, with only 2 of 7 cases that were still viable in the second trimester being identified in this way. Detection of nonhydropic Turner syndrome in the second trimester is more likely using the trisomy 18 algorithm,18 because triple marker levels are reduced in association with both aneuploidies.
There were too few cases of triploidy in this study to comment on the value of serum markers. Three of the eight triploidy cases were identified in the first trimester with cystic hygroma, and a further three miscarried, leaving only two cases with serum marker data in the first and second trimesters. We are therefore limited in the conclusions we can draw in this setting. Indeed, both the lethality and low prevalence of trisomy 13 and triploidy make these conditions unsuitable targets for screening.
Again, differential detection of triploidy by second-trimester Down syndrome and trisomy 18 algorithms is expected, based on whether the triploidy is maternally or paternally derived.21 The first-trimester serum marker pattern in cases of triploidy remains largely unknown.
In conclusion, the FASTER study data demonstrates that 78% of pregnancies complicated by trisomies 18 or 13, Turner syndrome, or triploidy would be identified through first-trimester Down syndrome screening using cystic hygroma, nuchal translucency ultrasonography, free β-hCG, and PAPP-A, and 69% would be detected using second-trimester quadruple screening, at 6.0% and 8.9% false-positive rates, respectively. This study confirms the Nicolaides et al22 first-trimester data, which identified an 88% sensitivity for trisomy 18 and trisomy 13 and 85% for other chromosomal defects at a 5% false-positive rate. Although first-trimester Down syndrome screening protocols can detect the majority of cases of non–Down syndrome aneuploidies, second-trimester protocols require trisomy 18 risk calculation rates to achieve high detection rates. There are insufficient numbers of these rarer aneuploidies to construct a robust individualized algorithm for each type of aneuploidy. Instead, application of existing Down syndrome and trisomy 18 screening algorithms can be used to detect the vast majority of these other aneuploid cases.
1. Wald NJ, Huttly WJ, Hackshaw AK. Antenatal screening for Down Syndrome with the quadruple test. Lancet 2003;361:835–6.
2. Malone FD, Canick JA, Ball RH, Nyberg DA, Comstock CH, Bukowski R, et al. First-trimester or second-trimester screening, or both, for Down's syndrome. N Engl J Med 2005;353:2001–11.
3. Jaques AM, Collins VR, Haynes K, Sheffield LJ, Francis I, Forbes R, et al. Using record linkage and manual follow-up to evaluate the Victorian maternal serum screening quadruple test for Down's syndrome, trisomy 18 and neural tube defects. J Med Screen 2006;13:8–13.
4. Benn PA, Fang M, Egan JF, Horne D, Collins R. Incorporation of inhibin-A in second-trimester screening for Down syndrome. Obstet Gynecol 2003;101:451–4.
5. 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.
6. Avgidou K, Papageorghiou A, Bindra R, Spencer K, Nicolaides KH. Prospective first-trimester screening for trisomy 21 in 30,564 pregnancies. Am J Obstet Gynecol 2005;192:1761–7.
7. Kagan KO, Avgidou K, Molina FS, Gajewska K, Nicolaides KH. Relation between increased fetal nuchal translucency thickness and chromosomal defects. Obstet Gynecol 2006;107:6–10.
8. Barkai G, Goldman B, Ries L, Chaki R, Zer T, Cuckle H. Expanding multiple marker screening for Down's syndrome to include Edward's syndrome. Prenat Diagn 1993;13:843–50.
9. Tul N, Spencer K, Noble P, Chan C, Nicolaides K. Screening for trisomy 18 by fetal nuchal translucency and maternal serum free beta-hCG and PAPP-A at 10–14 weeks of gestation. Prenat Diagn 1999;19:1035–42.
10. Benn PA, Leo MV, Rodis JF, Beazoglou T, Collins R, Horne D. Maternal serum screening for fetal trisomy 18: a comparison of fixed cutoff and patient-specific risk protocols. Obstet Gynecol 1999;93:707–11.
11. Spencer K, Nicolaides KH. A first trimester trisomy 13/ trisomy 18 risk algorithm combining fetal nuchal translucency thickness, maternal serum free beta-hCG and PAPP-A. Prenat Diagn 2002;22:877–9.
12. Hadlock FP, Shah YP, Kanon DJ, Lindsey JV. Fetal crown-rump length: reevaluation of relation to menstrual age (5–18 weeks) with high-resolution real-time US. Radiology 1992;182:501–5.
13. Malone FD, Ball RH, Nyberg DA, Comstock CH, Saade GR, Berkowitz RL, et al. First-trimester septated cystic hygroma; prevalence, natural history, and pediatric outcome. Obstet Gynecol 2005;106:288–94.
14. Malone FD, D'Alton ME, Society for Maternal-Fetal Medicine. First-trimester sonographic screening for Down syndrome. Obstet Gynecol 2003;102:1066–79.
15. Morris JK, Mutton DE, Alberman E. Revised estimates of the maternal age specific live birth prevalence of Down's syndrome. J Med Screen 2002;9:2–6.
16. Cuckle HS, Arbuzova S. Multimarker maternal serum screening for chromosomal abnormalities. In: Milunsky A, editor. Genetic disorders and the fetus: diagnosis, prevention, and treatment. 5th edition. Baltimore (MD): Johns Hopkins University Press; 2004. p. 795–835.
17. Palomaki GE, Haddow JE, Knight GJ, Wald NJ, Kennard A, Canick JA, et al. Risk-based prenatal screening for trisomy 18 using alpha-fetoprotein, unconjugated oestriol and human chorionic gonadotrophin. Prenat Diagn 1995;15:713–23.
18. Lambert-Messerlian GM, Saller DN Jr, Tumber MB, French CA, Peterson CJ, Canick JA. Second-trimester maternal serum inhibin A levels in fetal trisomy 18 and Turner syndrome with and without hydrops. Prenat Diagn 1998;18:1061–7.
19. Saller DN Jr, Canick JA, Schwartz S, Blitzer MG. Multiple-marker screening in pregnancies with hydropic and non-hydropic Turner syndrome. Am J Obstet Gynecol 1992;167:1021–4.
20. Benn PA, Ying J. Preliminary estimate for the second-trimester maternal serum screening detection rate of the 45,X karyotype using alpha-fetoprotein, unconjugated estriol and human chorionic gonadotropin. J Matern Fetal Neonatal Med 2004;15:160–6.
21. Benn PA, Gainey A, Ingardia CJ, Rodis JF, Egan JF. Second trimester maternal serum analytes in triploid pregnancies: correlation with phenotype and sex chromosome complement. Prenat Diagn 2001;21:680–6.
22. Nicolaides KH, Spencer K, Avgidou K, Faiola S, Falcon O. Multicenter study of first-trimester screening for trisomy 21 in 75 821 pregnancies: results and estimation of the potential impact of individual risk-orientated two-stage first-trimester screening. Ultrasound Obstet Gynecol 2005;25:221–6.