A model for second-trimester Down syndrome sonographic screening based on facial landmarks and digit length measurement.

Abstract

To determine whether Down syndrome can be detected by combining measurements of fetal nasal bone (NB) length, prenasal thickness (PT) and digits 2 and 3 of the hand. Two hundred and fifty-four normal and 25 Down-syndrome fetuses were scanned between 15 and 33 weeks' gestation. Physicians performing the scans were not blinded to the fetal karyotype. Both PT and NB were measured in a mid-sagittal plane. For PT measurement calipers were placed between the frontonasal angle and the outer skin edge. Digits 2 and 3 of one hand were also measured. The results (except for PT/NB ratio) were expressed in multiples of the normal gestation-specific median (MoM). A logistic regression model was used to estimate the odds of the fetus having Down syndrome given different combinations of NB, PT, PT/NB ratio, and digits 2 and 3 measurements. The odds were used to calculate the risk of Down syndrome for each pregnant woman from her age and measurements. The median PT MoM for unaffected fetuses and Down-syndrome fetuses was 1.12 vs. 1.35 (P < 0.0001). The median NB MoM for unaffected and Down-syndrome fetuses was 1.03 vs. 0.81 (P < 0.001) and the PT/NB ratio MoM for unaffected and Down-syndrome fetuses was 0.63 vs. 0.96 (P < 0.001). The respective median MoM values for digits 2 and 3 of the Down-syndrome fetuses were significantly smaller (0.81 vs. 0.93 and 0.89 vs. 0.95, respectively, P = 0.003). Only the PT/NB ratio and digit 2 were finally included in the logistic regression equation. Using a 1 in 200 risk cut-off, the observed sensitivity and false-positive rate were 76% and 6.7%, respectively. Combining the PT/NB ratio and digit 2 measurements yielded a promising screening detection rate. Confirmation of our findings in a prospective study is needed before the method can be used clinically.