Screening for Down's syndrome

Abstract

Just as screening for HIV infection in pregnancy is less than hoped for, so is screening for Down's syndrome. C. Ford and colleagues (pages 855–859) performed a survey of serum screening for Down's syndrome in a deprived population with a large proportion of women of Asian origin. Over five years more than sixteen thousand women were offered serum screening, of whom about three-quarters accepted the test and of those who were positive about half agreed to amniocentesis. One in twelve of the Asian women had positive serum tests compared with one in twenty of the Caucasian women, this difference being due to differences in the normal levels of hCG between the two groups of women; despite this only one-third of the Asian women agreed to amniocentesis, compared with two-thirds of the Caucasian women. The authors wonder if there are cultural pressures against amniocentesis in Asian women. There were nineteen cases of Down's syndrome during this time, of which fifteen resulted in a live birth. Four women underwent termination of pregnancy for Down's syndrome, and this was at a cost of four miscarriages in association with amniocentesis where serum screening had been positive. The authors conclude that the benefits to public health from serum screening in this population are small, and question the ethics and economics of continuing with the programme. Ford and colleagues point out however that serum screening was intrinsically successful in their study, for it was positive in eleven out of the thirteen cases of Down's syndrome in which testing was carried out; it was the low acceptance of amniocentesis which was responsible for the poor results of the screening programme. What if the screening procedures were changed to the first trimester with more reliance on ultrasound? Our knowledge of early screening tests is imperfect, with early enthusiasm being tempered by later caution. The hopes of many investigators have rested on fetal nuchal translucency, but B. J. Whitlow and colleagues (pages 872–876) show that this measurement depends very much on the attitude of the fetus: the average when the neck is flexed was 1.0 mm, in a neutral position 1.4 mm, and when the neck was extended 2.0 mm. The effects of these differences are enormous. In the 196 pregnancies in the study none of the nuchal translucency measurements would be above the 95th centile where the fetal neck is flexed, compared with 1.5% of measurements where the neck is in a neutral position, and 22% where the neck is extended. The authors recommend that nuchal translucency measurements are performed in the neutral position, since in that position the variation in measurement by a single observer is least. Can biochemical refinements be added to nuchal translucency measurements in the first trimester? Yes, say Roberto Biagiotti and his colleagues (pages 917–922), who measured pregnancy-associated plasma protein A and free β hCG in conjunction with nuchal translucency in 32 pregnancies with Down's syndrome compared with 200 normal pregnancies. With a false positive rate of 5% the combined tests detected three-quarters of pregnancies with Down's syndrome, compared with two-thirds using nuchal translucency alone. The authors acknowledge the limitations of their study with its small size and the selection of their cases, and recommend that larger studies of combined screening in unselected populations are carried out. A pregnancy may overcome the hurdles of screening for Down's syndrome in the first trimester by nuchal translucency and in the second trimester by serum screening, but how is one to interpret subtle abnormalities of the fetus found at a routine ultrasound scan in the second trimester which are associated with Down's syndrome and other aneuploidies? Leye Thompson and Baskaran Thilaganathan (pages 860–864) have investigated the significance of isolated hydronephrosis in 423 pregnancies with previous screening by nuchal translucency and serum biochemistry, and found no case of Down's syndrome. With more than one subtle abnormality the risk of Down's syndrome was twice that of the background risk, and the risk of other aneuploidies nine times the background risk. These studies illustrate the difficulties of screening for Down's syndrome. In some populations the acceptance of serum screening in the second trimester and subsequent amniocentesis may be so low as to question the justification of the screening programme, yet research into screening in the first trimester is confounded by small studies in selected pregnancies, which although giving useful information with which to generate hypotheses, cannot provide definite answers. What is required is a large study of screening for Down's syndrome in the first trimester in an unselected population, where not only is the efficiency of the tests measured, but also the women's perceptions of the screening.