A prenatal trisomy 21 screening program using alpha-fetoprotein, human chorionic gonadotropin, and free estriol...

The feasibility of large-scale Down syndrome maternal screening with dried blood samples and nonradioactive methods was examined.

A prospective observation study was performed on a nonselected population of 11,241 pregnant women sampled between January 1991 and September 1992, between 14 and 24 weeks' gestation (ultrasonographic scanning available for 91.6%), through a multicenter collaborative network. Enzyme-linked immunosorbent assays for alpha-fetoprotein, human chorionic gonadotropin, and free estriol were performed on dried blood samples. Risk determination was made with an in-house software implementing the multivariate gaussian log likelihood method.

A total of 10,450 samples were eligible for the study. Mean age at term was 27.9 years. A total of 6.84% of the patients were > or = 35 years old with a prior risk of trisomy 21 > 1:350. The general positive rate of our sample was 8.15%. After calculation 31.7% with prior risk > 1:350 were still in the high-risk group; 6.36% of the low-risk group were found to be at high risk for Down syndrome. Fifteen trisomic pregnancies were observed, of which 11 had a calculated risk higher than the selected cutoff value (1:350). The overall detection rate was 73%, specificity was 92%, and positive predictive power was 1.2%.

Our pilot study has shown performances within the range of conventional serum screening programs. Dried blood assays are a handy alternative to serum assays. Blot paper cards represent a simple method of sampling, well fitted for large population screening. Combined with nonradioactive methods, this method appears to be both low cost and effective. The current work apparently is the first large-scale Down screening program performed with dried blood.