Volume 199, Issue 3 , Pages 213-214, September 2008
Nuchal fold and nasal bone: how should we use them in Down syndrome screening?
Article Outline
Langdon Down in 1866 described many of the characteristics of children with the syndrome that now bears his name, but not until 1959 did the French human geneticist Jérôme Jean Louis Marie Lejeune discover the diagnostic test that would demonstrate the underlying chromosomal aberration. Hence, for almost 100 years, the diagnosis of Down syndrome could be suspected by a constellation of nondiagnostic physical findings but could not be confirmed. Even today, a pediatrician cannot be certain of the diagnosis of Down syndrome based on phenotype alone. Because many of the characteristic findings of trisomy 21 are also present in unaffected infants, a suspicion of Down syndrome is confirmed only by karyotype in about two-thirds of cases.1 This situation is analogous to antenatal screening in which certain characteristics of the fetus may suggest an affected pregnancy, but none are diagnostic. Odibo et al2 in this issue of the American Journal of Obstetricians and Gynecologists give us additional information that can be used to fine tune this evaluation.2
See related article, page 281
In their present study, Odibo et al2 compared the performance of nuchal fold (NF) thickness and nasal bone hypoplasia in a single population.2 As an independent marker, an absent nasal bone proved to be superior to an enlarged NF (>6 mm), with a likelihood ratio of 29 compared with 12. Based on these results, the authors rightly suggest that an absent nasal bone is the best single second-trimester ultrasound parameter for suspecting trisomy 21 and has the potential to identify about one-third of affected pregnancies. Because absent nasal bone and a very large NF are rare in a euploid population and their presence or absence can be identified relatively easily by ultrasound, screening with these dichotomous variables may be reasonable in centers with limited ultrasound experience. In such cases, the presence of either finding should suggest referral for further evaluation and possible invasive testing. However, in experienced centers conversion of these variables to likelihood ratios and incorporation of the results into more sophisticated Down syndrome risk algorithms is preferable.
Measurement of the NF in the second trimester was 1 of the earliest ultrasound markers that was used to alter Down syndrome risk and remains 1 of the most specific. Many studies, including that of Odibo et al, have used this measurement as a dichotomous variable with an upper limit threshold of >5 or >6 mm. With the use of the NF alone, the detection rate for Down syndrome is relatively unimpressive, ranging from 4-35%,3 and was only 12% in the present evaluation.2 Using NF in this way is suboptimal for 2 reasons. First, the NF thickness in unaffected pregnancies increases with advancing gestation, so a relatively small thickening could be more important at an earlier gestation than could a larger thickening later. For similar reasons, the use of a single cut-off with a fixed likelihood ratio throughout the entire second trimester is equally misleading. Second, the reassuring value of a small NF is diminished because women with moderately increased NFs are not treated differently than those with NF values near or below the median.
Because NF measurements fit a log Gaussian distribution, measurement-specific likelihood ratios can be calculated, for example, with the parameters of Borrell et al4 that are based on results in multiples of the median (MoM) for gestation. The differences in individual risk assessment compared with a rigid cut off are significant. With a standard 6-mm cut-off that is associated with a likelihood ratio of 12, all women with such measurements would be screen positive, with even the youngest women having a risk of approximately 1:100. Alternatively, with the use of NF measurements as a continuous variable and with their conversion to gestational age–specific likelihood ratios, a woman with an a priori risk of 1:1000 and a 6-mm NF at 20 weeks of gestation (biparietal diameter [BPD], 49; MoM, 1.6) has an adjusted Down syndrome risk of 1:410. The same women at 14 weeks (BPD, 30; MoM, 2.7) has a risk of 1:60. These individual risk assessments are even more accurate when other variables such as biochemical analytes or other ultrasound markers are added.
More recently nasal bone hypoplasia has been suggested as a similarly powerful marker in the second trimester.5 When second-trimester ultrasound imaging is used, Odibo et al2 demonstrated complete absence of ossification in approximately 30% of affected pregnancies and 1% of chromosomally normal fetuses. Nasal bone hypoplasia, defined as <0.75 MoM, identified 47% of Down syndrome pregnancies and occurred in 6% of unaffected pregnancies. Nasal bone measurements, as with NF, appear to fit a log Gaussian distribution so that measurement-specific likelihood ratios can be calculated rather than the use of a fixed cut-off,6 which should improve individual risk assessment when combined with maternal age and other variables.
A number of concerns must be addressed before the introduction of nasal bone measurements into Down syndrome screening can be accepted. First, the number of affected pregnancies in the available reports is limited, and prospective blinded evaluations of this approach are few. Second, the 5 reports of nasal bone measurements that are available show differences in their gestational age–specific median equations, which emphasizes the need for standardization of the measurement technique and development of a quality monitoring approach, such as that used in the first trimester. Third, the size of the normal nasal bone may vary by ethnicity, which would require that either ethnic group–specific curves or a standard correction factor for each group be developed. Despite these limitations and whether one uses a single cut-off or a likelihood ratio, the integration of the nasal bone into a risk algorithm that includes other markers (either ultrasound or serum analyte) appears to maximize the performance of second-trimester screening and results in each patient getting their most accurate individual risk assessment.
Maximization of the accuracy of each patient's individual Down syndrome risk is of ever increasing importance. Maternal age is no longer an independent indication for invasive testing, so women >35 years old depend on this evaluation in deciding whether to undertake the risk of a diagnostic procedure. Likewise, with the recommendation of the American College of Obstetricians and Gynecologists that all women should be offered the opportunity of invasive testing, these women could use their individual risk assessment results to guide their decision-making.
References
- . Accuracy of clinical diagnosis in Down's syndrome. Arch Dis Child. 2004;89:691
- . Comparing the efficiency of second-trimester nasal bone hypoplasia and increased nuchal fold in Down syndrome screening. Am J Obstet Gynecol. 2008;199:281.e1–281.e5
- Second-trimester ultrasound to detect fetuses with Down syndrome a meta-analysis. JAMA. 2001;285:1044–1055
- . Criteria for fetal nuchal thickness cut-off: a re-evaluation. Prenatal Diagn. 1997;7:23–29
- . Nasal bone assessment in prenatal screening for trisomy 21. Am J Obstet Gynecol. 2006;195:1219–1223
- . A model for second-trimester Down syndrome sonographic screening based on facial landmarks and digit length measurement. Ultrasound Obstet Gynecol. 2006;27:290–295
PII: S0002-9378(08)00709-6
doi:10.1016/j.ajog.2008.06.079
© 2008 Mosby, Inc. All rights reserved.
Refers to article:
-
Comparison of the efficiency of second-trimester nasal bone hypoplasia and increased nuchal fold in Down syndrome screening
Volume 199, Issue 3 , Pages 213-214, September 2008
