Current indications for cytogenetic evaluation leave the majority of Down syndrome fetuses undetected. Using advanced maternal age and low maternal serum alpha-fetoprotein (AFP) levels as criteria, only 40% of fetuses with Down syndrome (trisomy 21) are identified (positive predictive value, 0.4% to 1%). We evaluate the sonographically detectable physical features of second trimester fetuses to determine whether these features are more sensitive and specific than maternal age for detecting fetuses with abnormal karyotypes. From March 1, 1990, to September 1, 1991, more than 5,000 fetuses between 14 and 20 weeks of development were referred for genetic amniocentesis because of advanced maternal age or abnormal AFP levels. Forty-three of these 5,000 fetuses were later found to have autosomal trisomies by karyotype (32 with trisomy 21, nine with trisomy 18, and two with trisomy 13). A sample of 588 consecutive normal fetuses from the total of more than 5,000 amniocenteses performed during this period of time was used as our control group for statistical analysis. The sonographic features of these 588 normal second trimester fetuses and the 43 trisomic fetuses recorded prospectively prior to knowledge of the karyotype were evaluated statistically. The femur and humerus lengths, nuchal fold, renal pelvic dimension, and major structural defects were compared in the normal and trisomic fetuses. On the basis of our results, a weighted sonographic score was developed to optimize the detection of fetuses at risk for aneuploidy. Using our previously published formulas and criteria for a short femur and humerus, 17/32 (53%) fetuses with Down syndrome and 23/588 (3.9%) of the normal fetuses were identified, Twenty two of 32 Down syndrome fetuses (69%) and 2/588 (0.34%) of normals had a nuchal fold greater-than-or-equal-to 6 mm, and 11 of 32 Down syndrome fetuses and all those with trisomies 18 and 13 had a major anomaly detected sonographically. The following scoring system was developed for the detection of aneuploidy: nuchal fold = 2, major structural defect = 2, and short femur, short humerus, and pyelectasis = 1 each. Selecting fetuses with a score of greater-than-or-equal-to 2 would identify 26/32 (81%) Down syndrome fetuses, and 9/9 (100%) and 2/2 (100%) fetuses with trisomies 18 and 13 respectively, but only 26/588 (4.4%) of the normal fetuses. Using the sonographic score of 2 results in a positive predictive value for a 1/250 risk group of 6.87% for identifying Down syndrome fetuses and 7.25% for all three trisomies. For the 1/500 and 1/1,000 risk groups, the positive predictive values are 3.75% and 1.91%, respectively, for identifying the three trisomies. Sonographic features suggesting aneuploidy appear more effective in identifying affected fetuses than advanced maternal age.
