First-trimester screening for fetal Down syndrome and other aneuploidies has been described using a combination of sonographic measurement of the nuchal translucency space and the measurement of the serum markers free β human chorionic gonadotropin (βhCG) and pregnancy associated plasma protein A (PAPP-A).1 This approach to risk assessment has been reported to detect approximately 82% of cases of Down syndrome, with a 5% false-positive rate.1 A major advantage of Down syndrome risk assessment during the first trimester of pregnancy may be the ability to provide information to patients earlier in gestation, thereby allowing prenatal diagnosis by means of chorionic villus sampling (CVS), with earlier reassurance for the majority of patients and the option for safer pregnancy termination.
Recent evidence suggests that evaluation of the fetal nasal bones might also help to identify fetuses at increased risk for aneuploidy. Specifically, absent or hypoplastic nasal bones has been suggested as a powerful marker for identification of fetal aneuploidy in both the first and second trimesters of pregnancy.2–4 In the original study describing this technique, Cicero and colleagues2 evaluated the fetal nasal bones at the time of nuchal translucency sonography at 11 to 14 weeks in 701 high-risk pregnancies immediately before CVS. Absence of the fetal nasal bones detected 73% of Down syndrome pregnancies, with a false-positive rate of only 0.5%. This same group of investigators has recently updated their experience with this technique to include 5,918 high-risk pregnancies undergoing CVS, and demonstrated a 69% Down syndrome detection rate for a 2.5% false-positive rate.3
Evaluation of the nasal bones during the first trimester has been limited to very specialized centers. Nearly all studies of the usefulness of this first trimester technique for detection of fetal aneuploidy have been derived from high-risk pregnancies undergoing CVS or from select specialist referral prenatal diagnosis centers.3,5–7 The only study derived from an unselected population is of limited usefulness, because pregnancy outcome was only available on 65% of enrolled patients.8 The performance of nasal bone sonography in the hands of select experts, or the determination of nasal bone absence in select high-risk pregnancies, will not allow for an objective assessment of the usefulness of this technique for general population screening. Before endorsing this new marker as a worthwhile screening tool for general population screening in the community, data are needed to confirm both the success rate at obtaining an adequate nasal bone image and the aneuploidy detection rate in this setting. The objective of this study, therefore, was to evaluate the performance of nasal bone sonography as a screening tool for fetal aneuploidy in an unselected general patient population by means of a prospective multicenter trial.
MATERIALS AND METHODS
The First And Second Trimester Evaluation of Risk (FASTER) Trial is a multicenter prospective study designed to compare different forms of screening for fetal aneuploidy. From 1999 until 2002, this trial enrolled 38,189 singleton pregnancies between 10 3/7 weeks and 13 6/7 weeks of gestation for nuchal translucency sonography at 15 centers throughout the United States. Enrolled patients also submitted first-trimester serum samples for assay of PAPP-A and free βhCG as well as second-trimester serum samples for assay of alpha fetoprotein, hCG, unconjugated estriol, and Inhibin-A. The study population included patients 16 years of age or older with viable singleton pregnancies and fetal crown–rump lengths between 36 mm and 79 mm. The presence of a septated cystic hygroma or an open neural tube defect at the time of the nuchal translucency sonographic evaluation were the only exclusion criteria, although all such excluded patients were followed up separately. All nuchal translucency ultrasound examinations were performed by specially trained sonographers and sonologists, who followed a uniform technique, submitted a minimum of 50 images confirming mastery of the technique, and were continuously evaluated by an ongoing quality control process.
After the emergence of initial studies suggesting a role for first-trimester nasal bone evaluation in 2001, the FASTER Trial research protocol was modified to add the evaluation of the fetal nasal bones at the time of the first-trimester nuchal translucency ultrasound examination. From January 2002 until April 2002, 45 of the specially trained nuchal translucency sonographers at 12 of the 15 FASTER enrollment centers underwent additional uniform training in first-trimester nasal bone sonography. From May 2002 until the completion of the FASTER Trial in December 2002, all patients presenting for nuchal translucency sonography at one of the 12 participating FASTER sites in which a certified nasal bone sonographer was available also underwent nasal bone sonographic evaluation. All sonographers followed a standard protocol for imaging the nasal bones, similar to that followed by Cicero and colleagues,2,3 which is summarized as follows: 1) fetus imaged in a perfect mid-sagittal plane with fetal spine down, 2) angle of insonation of ultrasound beam with fetal profile close to 45o, 3) image magnified significantly until 2 echogenic lines visible in region of fetal nose, 4) transducer tilted from side-to-side to distinguish fetal skin from underlying nasal bones, and 5) deeper echogenic line noted to become more echolucent at its distal end. The sonographer then recorded the nasal bones as being present, absent, or unable to determine. If adequate imaging could not be obtained using transabdominal sonography, transvaginal sonography was used before considering the nasal bones indeterminate.
Nasal bone sonography quality control was confirmed as follows: After completion of standardized training, study sonographers submitted a minimum of 10 images for central review to confirm adequate nasal bone sonography. A second study sonographer, also certified in nasal bone sonography, was required to evaluate each nasal bone image before the patient was discharged. Images were only considered acceptable if each of the components of the sonographic protocol were followed. Up to 20 minutes were allotted for completion of nuchal translucency and nasal bone evaluation before the sonographers were allowed to consider the nasal bones as indeterminate. All ultrasound machines used during this study allowed measurements to one tenth of a millimeter, and all were recalibrated every 3 months.
Information on presence or absence of fetal nasal bones was not used clinically during this trial and was not provided to either patients or referring physicians. After the completion of first-trimester and second-trimester screening, patients were provided with individualized Down syndrome risk assessment based on maternal age, nuchal translucency, PAPP-A, free βhCG, alpha fetoprotein, hCG, unconjugated estriol, and Inhibin-A. Separate first- and second-trimester Down syndrome risk assessments were provided, and patients who were screen positive from either of these assessments were offered amniocentesis. Screen-positive patients who declined amniocentesis and those who had a miscarriage were approached for neonatal cord blood or fetal tissue karyotyping. All enrolled patients were followed up for complete pregnancy and pediatric outcome. A software tracking program, with up to 10 contact options per patient, was used to ensure optimal outcome ascertainment. All patients with positive screening results or with any abnormal pregnancy or pediatric outcome event also had photocopies of their medical records and those of their infants sent for central medical review. An additional 10% random sample of all other medical records was also submitted for central review.
Descriptive statistics were generated for all study variables, including means, standard deviations, medians, and interquartile ranges for continuous variables and relative frequencies for categorical variables. The proportions of cases in which an acceptable nasal image was obtained were estimated, together with 95% confidence intervals. Crude odds ratios were estimated using logistic regression analysis. Maternal characteristics that were significant were entered into a multiple logistic regression model to simultaneously assess the association between characteristics and successful imaging. Performance characteristics of nasal bone sonography with respect to detection of aneuploidy, including sensitivity, specificity, and positive predictive value were estimated. All analyses were conducted in SAS 6.12 (SAS Institute Inc., Cary, NC). This study was reviewed and approved by the Institutional Review Boards of all 12 participating centers.
A total of 42,367 patients were evaluated for possible enrollment in the FASTER Trial at 15 clinical centers in the United States from October 1999 until December 2002. Of these patients, 4,178 were found to be ineligible for enrollment, mostly because of incorrect gestational age (2,636 cases), multiple gestation (896 cases), or age less than 16 years (357 cases). Of 38,189 eligible patients who underwent first-trimester screening with nuchal translucency sonography, 6,324 were enrolled at one of the 12 sites performing nasal bone sonography during the period relevant to this study of May to December 2002.
Of 6,324 patients enrolled into the nasal bone sonography study, complete pregnancy and pediatric outcome ascertainment was obtained in 6,228, or 98.5% of cases. Karyotype information was available in 587 cases, including 510 from amniocentesis, 41 from neonatal cord blood, and 36 from products of conception and autopsy material. This represented 17% of patients who had a pregnancy loss or termination, and 10% of screen-positive patients who declined amniocentesis. Mean maternal age of enrolled patients was 30.1 years (standard deviation 5.7, range 16–47), and 22.1% were 35 years old or older at the expected date of delivery. The maternal race distribution included 64.8% white, 25.8% Hispanic, 4.6% African American, 3.9% Asian, and 0.9% other. These demographic characteristics were not statistically significantly different from those of the entire FASTER cohort of 38,189 cases. A total of 22 cases had absent nasal bones, and the race distribution of these cases was 77.3% white, 13.6% Hispanic, 4.5% African American, and 4.5% other.
First-trimester sonography successfully provided adequate nasal bone imaging in 4,801 cases (75.9%, 95% confidence interval [CI] 74.4–76.8), but was unsuccessful at allowing adequate nasal bone imaging in 1,523 cases (24.1%, 95% CI 22.0–26.2). The second certified study sonographer disagreed with the initial sonographers’ nasal bone determination in only 89 (2%) cases. Predictive factors of failure to achieve adequate nasal bone visualization are summarized in Table 1. Failure to obtain adequate nasal bone imaging was significantly more likely at 10 weeks and 13 weeks of gestation, compared with 11 weeks and 12 weeks of gestation. The mean maternal body mass index (measured during the first trimester) was significantly greater in patients with failed nasal bone imaging compared with patients with successful nasal bone imaging. Nasal bone imaging was significantly more likely to fail when nuchal translucency sonography was technically unsuccessful. Use of transvaginal sonography was also significantly more likely to be associated with failure of nasal bone imaging.
Figure 1 describes the pregnancy outcome of the 6,324 patients in which first-trimester nasal bone sonography was attempted. Only 22 cases (0.5%, 95% CI 0.3–0.7%) with successful imaging were described as having absent nasal bones. Sonographic absence of the nasal bones in the first trimester detected none of the 11 cases of Down syndrome in this population, with 9 of the 11 cases described as having nasal bones present and 2 of the 11 cases described as being indeterminate. The only other aneuploidies were 2 cases of trisomy 18, with 1 each described as having nasal bones present and 1 as absent. The overall sensitivity of first trimester absence of the nasal bones for fetal aneuploidy was 7.7% (1/13) (95% CI 0.2–36.0%), with a false-positive rate of 0.3% (95% CI 0.2–0.5%). The positive predictive value of absent nasal bones for fetal aneuploidy was 4.5% (1/22) (95% CI 0.1–22.8%).
Complete pregnancy and pediatric outcome was obtained on 98.5% of all enrolled patients and additionally on all 22 cases with absent fetal nasal bones. In this group of 22 cases, the only abnormalities found on detailed pediatric outcome evaluation were 1 case of trisomy 18, 1 case of transient neonatal supraventricular tachycardia, and 1 atrial septal defect. No other chromosomal abnormalities were present.
After the results of this form of aneuploidy screening were calculated, further data were obtained on the individual sonographers who performed the nasal bone sonography in each of the 13 cases of fetal aneuploidy. These 13 nasal bone evaluations were performed by 13 different sonographers at 6 different medical centers. Table 2 summarizes the number of nuchal translucency ultrasound examinations completed by each of these 13 sonographers before evaluating their aneuploidy case. The vast majority of sonographers involved were highly experienced first-trimester sonographers, with a median of 515 nuchal translucency ultrasound examinations performed before their index aneuploidy case, and a median success rate at obtaining nuchal translucency measurements of 99%. All except 2 sonographers had completed more than 200 nuchal translucency ultrasound examinations, and several had already completed more than 1,000 examinations before their index aneuploidy case.
With the increasing popularity of first-trimester risk assessment for fetal aneuploidy, new sonographic markers are being sought that may improve the performance of nuchal translucency–based screening programs. The FASTER Trial provided an ideal forum in which to test the hypothesis that routine assessment of nasal bones during the first trimester is a useful screening tool in an unselected general population. Our results derive from one of the largest studies of nasal bone sonography published to date and is notable as being derived from an unselected population in which nearly complete outcome ascertainment was obtained. Our results indicate that evaluation of the nasal bones during the first trimester does not improve detection of aneuploidy in an unselected population.
A group of investigators from the Fetal Medicine Foundation in London has suggested that first-trimester absence of the nasal bones may detect 69% of cases of Down syndrome, with a 2.5% false-positive rate. However, these results were derived exclusively from a high-risk cohort who were undergoing CVS.2,3 Four other studies have also been published evaluating the role of first-trimester nasal bone absence and fetal aneuploidy. Otano and colleagues5 evaluated 194 high-risk pregnancies for fetal nasal bones at 11 to 14 weeks of gestation immediately before CVS. They noted absence of the nasal bones in 3 of the 6 (50%) Down syndrome cases compared with 1 of 184 (0.5%) normal pregnancies. Orlandi and colleagues6 studied 1,089 pregnancies between 11 and 14 weeks of gestation and noted absence of the fetal nasal bones in 10 of 15 (67%) Down syndrome pregnancies compared with 12 of 1,002 (1.2%) normal pregnancies. This study was performed at 3 specialist referral prenatal diagnosis centers, and many of the patients were enrolled just before CVS. Viora and colleagues7 evaluated the fetal nasal bones in 1,906 patients at 11 to 14 weeks of gestation and found nasal bone absence in 6 of 10 (60%) Down syndrome cases and 24 of 1,733 (1.4%) normal pregnancies. This study was also performed at a specialist referral prenatal diagnosis center, and many of the patients were being evaluated immediately before CVS. Finally, Zoppi and colleagues8 studied the fetal nasal bones in 5,532 fetuses from 11 to 14 weeks of gestation and noted nasal bone absence in 19 of 27 (70%) Down syndrome fetuses, compared with 7 of 3,463 normal fetuses. This last study was compromised by the fact that pregnancy outcome information was available on only 65% of enrolled patients, thereby calling into question the true ascertainment of aneuploid pregnancies. It should be noted, however, that not all investigators have found first-trimester absence of fetal nasal bones to be associated with Down syndrome. De Biasio and Venturini9 described a series of 5 consecutive cases of Down syndrome in which fetal nasal bones were considered to be present.
The first conclusion derived from our study is that adequate nasal bone imaging could not be obtained in one quarter (24%) of all cases. This finding was despite careful additional training of select, highly experienced, first-trimester sonographers, who all followed a uniform sonographic technique. Additionally, our study is the only one described to date in which a quality control program to monitor ongoing performance of nasal bone sonography was in place. This high failure rate was surprising, particularly as it is in contrast to the 5 prior studies published on this technique, in which failure rates ranged from 0.1% to 8.1%.3,5–8 The reasons for our high failure rate are uncertain. It is possible that when sonographers evaluate a subjective sonographic marker, such as the presence or absence of the nasal bones, in a low-risk patient population they may be less compulsive or precise in their technique compared with evaluation of a patient known to be at increased risk for aneuploidy. Additionally, in a research trial in which nasal bone sonography results are not used clinically, sonographers may also not be as precise in their technique. It is possible that higher success rates at adequate nasal bone imaging may be obtained if unlimited time is allowed, although this may not be practical in a busy clinical practice. Finally, it is possible that our criteria for judging adequacy of nasal bone images may have been more stringent than prior studies, although none of the prior studies provide data on quality assurance methods that were used.
We believe that our results represent a realistic assessment of the likely performance of this technique should it be implemented widely in general practice. If sonographers who have completed 1,000 to 2,000 nuchal translucency ultrasound examinations, in the highly supervised setting of a clinical research trial, cannot reliably evaluate the nasal bone in the first trimester, then this does not bode well for the widespread implementation of this technique in less experienced or less regulated settings. It should also be noted that 10 of the 13 cases of aneuploidy in our study were found in patients who had successful nasal bone sonography completed. Therefore, even if the 2 aneuploidy cases in the failed nasal bone sonography group were detected, it would not have significantly altered our conclusions.
In our experience, nasal bone sonography is more difficult to complete than nuchal translucency sonography, because imaging of the nasal bones requires a perfect mid-sagittal image and optimal angle of insonation with the fetal profile, whereas nuchal translucency measurements can still be obtained with minor variations off-center and differences in direction of imaging. Demonstrating the absence of a very small structure is even more difficult than detecting its presence, because it can be difficult to know for certain whether the nasal bones are absent or whether the images are simply suboptimal. Therefore, application of nasal bone sonography will be limited at early gestational ages when the nasal bones are particularly small, as seen in our study by the 31% failure rate at 10 weeks of gestation.
Other factors associated with an increased failure rate of nasal bone imaging in our study included larger maternal body habitus, inadequate nuchal translucency sonography, and use of a transvaginal sonographic approach. The finding of increased body mass index among patients with failure of nasal bone imaging is not surprising, because obesity is a recognized practical limitation of ultrasonography. It is also not surprising that sonographers were more likely to fail to obtain adequate nasal bone imaging if they had already failed to obtain an adequate nuchal translucency image, because the 2 forms of screening share similar approaches. Similarly, given that our protocol called for use of the transvaginal probe if adequate imaging was not obtained with the transabdominal approach, it is not surprising that we found higher failure rates when the transvaginal probe was needed.
The second major conclusion from this trial is that absence of the fetal nasal bones is a poor screening tool for fetal aneuploidy in the general, unselected population. None of the 11 cases of Down syndrome were detected using this technique, and only 1 of the 2 trisomy 18 cases was detected. These findings are also in contrast to the 5 prior publications on this technique, in which Down syndrome detection rates of 50% to 70% were described.3,5–8 However, because our study is the largest published to date, and the only study derived from an entirely unselected population with adequate ascertainment, we believe that our results more accurately reflect what would be achieved by this form of screening if implemented on a national scale. Our findings do not dismiss the results of the 5 prior studies, because they clearly suggest that there may be individuals in select referral centers who may have the ability to detect a number of aneuploidy cases using first-trimester nasal bone sonography.3,5–8 However, our study does not support extrapolation of these results to the general population, and thus first-trimester nasal bone sonography cannot be recommended as a screening tool on a national basis.
In conclusion, we feel that this study suggests that there is no role for this form of screening for fetal aneuploidy in the general population. In the hands of highly trained and highly experienced sonographers, adequate imaging of the nasal bones could not be obtained in one quarter of cases. Although there may be select individuals and centers that can identify a number of fetuses with Down syndrome based on the absence of nasal bones at first-trimester sonography, this is insufficient reason to recommend the widespread implementation of this screening modality in the general population. The additional burden associated with its widespread use would be difficult to justify, particularly when more reliable screening methods are available. On the contrary, our data would suggest that the majority of aneuploidy cases would be missed using this technique.
1. Malone FD, D'Alton ME, for the Society for Maternal Fetal Medicine. First-trimester sonographic screening for Down syndrome. Obstet Gynecol 2003;102:1066–79.
2. Cicero S, Curcio P, Papageorghiou A, Sonek J, Nicolaides K. Absence of nasal bone in fetuses with trisomy 21 at 11–14 weeks of gestation: an observational study. Lancet 2001;358:1665–7.
3. Cicero S, Rembouskos G, Vandecruys H, Hogg M, Nicolaides KH. Likelihood ratio for trisomy 21 in fetuses with absent nasal bone at the 11-14-week scan. Ultrasound Obstet Gynecol 2004;23:218–23.
4. Bromley B, Lieberman E, Shipp TD, Benacerraf BR. Fetal nose bone length: a marker for Down syndrome in the second trimester [published erratum appears in J Ultrasound Med. 2003:162]. J Ultrasound Med 2002;21:1387–94.
5. Otano L, Aiello H, Igarzabal L, Matayoshi T, Gadow EC. Association between first trimester absence of fetal nasal bone on ultrasound and Down syndrome. Prenat Diagn 2002;22:930–2.
6. Orlandi F, Bilardo CM, Campogrande M, Krantz D, Hallahan T, Rossi C, et al. Measurement of nasal bone length at 11–14 weeks of pregnancy and its potential role in Down syndrome risk assessment. Ultrasound Obstet Gynecol 2003;22:36–9.
7. Viora E, Masturzo B, Errange G, Sciarrone A, Bastonero S, Campogrande M. Ultrasound evaluation of fetal nasal bone at 11 to 14 weeks in a consecutive series of 1906 fetuses. Prenat Diagn 2003;23:784–7.
8. Zoppi MA, Ibba RM, Axiana C, Floris M, Manca F, Monni G. Absence of fetal nasal bone and aneuploidies at first-trimester nuchal translucency screening in unselected pregnancies. Prenat Diagn 2003;23:496–500.
9. De Biasio P, Venturini PL. Absence of nasal bone and detection of trisomy 21. Lancet 2002;359:1344.
The other members of the FASTER Research Consortium that contributed to this study are: K. Brigatti, ms, R. Denchy, ms (Columbia University, NY, NY), F. Porter, md, L. Cannon, bs, K. Nelson, bsn, C. Loucks, rnc, A. Yoshimura (University of Utah, and IHC Perinatal Centers, Salt Lake City, Provo, and Ogden, UT), D. Luthy, md, S. Coe, ms (Swedish Medical Center, Seattle, WA), J. Esler, bs (William Beaumont Hospital, Royal Oak, MI), G. Hankins, md, R. Bukowski, MD, J. Lee ms, (UTMB Galveston, TX) K. Eddleman, md, Y. Kharbutli ms (Mount Sinai Medical Center, NY, NY), J. Hobbins, md, L. Schultz, rn (University of Colorado Health Science Center, Denver, CO), D. Bianchi, md, B. Isquith, ms, B MacKinnon, rn, (Tufts University, Boston, MA), J. Canick, phd, G. Messerlian, phd, C. Duquette, rdms (Brown University, Providence, RI), R. Baughman, ms (University of North Carolina, Chapel Hill, NC), D. Alexander, md, J. Hanson, md, F. de la Cruz, md (National Institute of Child Health and Human Development) K. Dukes, phd, D. Emig, mph, J. Vidaver, ma (DM-STAT, Inc, Medford, MA).