Ratio of nuchal thickness to humerus length for Down syndrome detection

OBJECTIVE: Ultrasonographic biometry markers are now being used clinically to adjust Down syndrome risk. The limitations are that the definitions of "abnormal" measurements used are arbitrary, thus reducing screening performance, and also that patient-specific Down syndrome risks cannot be calculated. We report a new ultrasonographic algorithm that is sensitive for Down syndrome detection and that estimates individual risk. STUDY DESIGN: Overall in fetal populations with Down syndrome the humerus length is decreased, whereas the nuchal thickness is increased relative to that of a normal population. The nuchal thickness/humerus length ratio therefore shows an even greater increase and magnifies the separation between Down syndrome and healthy groups. Prospective data were collected in midtrimester amniocentesis cases. A regression equation for the median nuchal thickness/humerus length ratio based on biparietal partial diameter was generated. The Down syndrome likelihood ratio, or the odds on the basis of the nuchal thickness/humerus length ratio (multiples of the median), was multiplied by the age-related risk to give the posterior Down syndrome risk. Charts for rapid estimation of individual Down syndrome risk on the basis of maternal age and the nuchal thickness/humerus length ratio were constructed. RESULTS: There were 94 cases of Down syndrome and 4700 cases in which the karyotype was normal. The mean (±SD) gestational age of the study population was 16.1 ± 1.6 weeks. Thirty-three fetuses with Down syndrome and 68 karyotypically normal fetuses had gross anomalies. The equation for the expected median nuchal thickness/humerus length ratio was as follows: 10_e(1.7163 - 0.0292) × BPD + 0.0003 × BPD², where BPD is the biparietal diameter. In the overall study population the nuchal thickness/humerus length ratio and maternal age had a 79.8% detection rate at a 22.1% false-positive rate, compared with maternal age plus humerus length (sensitivity, 55.1%) or maternal age plus nuchal thickness (sensitivity, 66.7%) at the same false-positive rate. For women ≥35 years old the values were 80% and 22.0%, respectively. CONCLUSIONS: We report an ultrasonographic biometry algorithm that, in combination with maternal age, detects 79.6% of Down syndrome cases in a high-risk group. Individual Down syndrome risk can be quickly calculated at the bedside and made available to women who desire this information before making a decision on amniocentesis. On the basis of published standards, ultrasonographic biometry as described would be a cost-effective alternative to amniocentesis in this high-risk group.