Karyotype-phenotype correlations of common trisomy mosaicism prenatally diagnosed via amniocentesis was reviewed in 305 new cases from a collaboration of North American cytogenetic laboratories. Abnormal outcome was noted in 10/25 (40%) cases of 47,+13/46, 17/31 (54%) cases of 47,+18/46, 10/152 (6.5%) cases of 47,+20/46, and in 49/97 (50%) cases of 47,+21/46 mosaicism. Risk of abnormal outcome in pregnancies with less than 50% trisomic cells and greater than 50% trisomic cells were: 26% (4/15) versus 60% (6/10) for 47,+13/46, 52% (11/21) versus 75% (6/8) for 47,+18/46, 4.5% (6/132) versus 20% (4/20) 47,+20/46, and 45% (27/60) versus 59% (22/37) for 47,+21/46. Phenotypically normal liveborns were observed with mean trisomic cell lines of 9.3% for 47,+13/46, 8.6% for 47,+18/46, 27% for 47, +20/46, and 17% for 47,+21/46. Cytogenetic confirmation rates were 46% (6/13 cases) for 47,+13/46 mosaicism, 66% (8/12 cases) for 47, +18/46, 10% (10/97 cases) for 47,+20/46, and 44% (24/54 cases) for 47,+21/46. There were higher confirmation rates in pregnancies with abnormal versus normal outcome: 50% versus 44% for 47,+13/46 mosaicism, 100% versus 33% for 47,+18/46, 66% versus 7% for 47, +20/46, and 55% versus 40% for 47,+21/46. Repeat amniocentesis is not helpful in predicting clinical outcome. It may be considered when there is insufficient number of cells or cultures to establish a diagnosis. Fetal blood sampling may have a role in mosaic trisomy 13, 18, and 21 as the risk for abnormal outcome increases with positive confirmation: 1/5 (20%) normal cases versus 5/8 (62%) abnormal cases. High resolution ultrasound examination(s) is recommended for clinical correlation and to facilitate genetic counselling.

Trisomy for an F‐like chromosome with banding characteristics of chromosome 20 was observed, respectively, in 25 and 8% of amniotic fluid cell clones in two cases of prenatal diagnosis for advanced maternal age. No such aneuploidy was found upon examination of fetal or postnatal tissues. In both instances there were no detectable malformations. It is suggested that propensity to nondisjunction for an F‐like chromosome, not unlike tetraploidy, might be another characteristic of amniotic fluid cells in culture.

The last decade has witnessed increasing application of human cytogenetic technology to prenatal chromosome analysis. However, unlike the rather uniform peripheral blood T-lymphocyte system which has provided most of our experience in human cytogenetics, long-term amniotic-fluid cell cultures display extreme cellular heterogeneity and disproportionate growth of certain cell types as a consequence of clonal amplification. When they enter cell culture, many of these cells are approaching the terminal stages of their respective, life spans and may have accumulated chromosomal aberrations. Concern about the possibility of true fetal mosaicism seems warranted chiefly in situations were multiple colonies display potentially viable aberrations. Clonal analysis, preferably of multiple clonal types, and attention to details of clonal morphology are likely to minimize diagnostic errors and undue apprehension resulting from mosaicism in amniotic-fluid cell cultures.

An aneuploid karyotype with an extra submetacentric C‐group chromosome was observed in all metaphase cells in 5 of 24 primary amniotic fluid cell clones and in admixture with normal cells in two additional clones. Trisomy 8 was demonstrated by R‐banding. The parents elected to terminate the pregnancy. Successful cultures were initiated from nine fetal tissues and aneuploid cells were observed in cultures derived from two separate sites in the skin.