Combining fetal nuchal fold thickness with second-trimester biochemistry to screen for trisomy 21

To assess second-trimester screening for trisomy 21 by combining ultrasound nuchal fold (NF) measurement with maternal serum biochemistry.

NF, maternal serum alpha-fetoprotein (AFP) and free beta-human chorionic gonadotropin (beta-hCG) were determined concurrently at 14-19 weeks' gestation in a study population comprising 1813 women with singleton pregnancies, including 1257 unselected women undergoing serum screening for trisomy 21 (1999-2002), and 556 high-risk pregnancies prior to amniocentesis (2003-2005), 402 of whom had positive serum screening tests. The results were expressed in multiples of the gestation-specific normal median (MoMs).

There were 1799 unaffected singleton pregnancies, and their NF values approximately fitted a log Gaussian distribution over a wide range. There was a weak but statistically significant correlation between log NF and log AFP (r = - 0.069, P < 0.005) and the correlation coefficient between log NF and log free beta-hCG was even smaller and not statistically significant (r = 0.038, P = 0.11). Among the seven trisomy 21 pregnancies, the median NF level was 1.53 MoM (geometric mean 1.75 MoM), a highly statistically significant increase compared with unaffected pregnancies (P < 0.0001, one-tail Wilcoxon Rank Sum Test). In pregnancies referred because of positive serum screening tests (391 unaffected, seven cases of trisomy 21, one of monosomy X and three other chromosomal anomalies) the use of NF to modify the serum screening risk would have reduced the invasive procedures in unaffected pregnancies by 46% without affecting the detection rate of trisomy 21 or other anomalies. Statistical modeling predicted that adding NF to AFP and free beta-hCG would increase detection more than would adding unconjugated estriol as well as inhibin-A, an analyte that is difficult to measure with precision.

The addition of NF measurement to second-trimester biochemical markers improves screening performance, and could overcome drawbacks in the implementation of inhibin-A assay in clinical practice.