First-trimester risk calculation for trisomy 13, 18, and 21: comparison of the screening efficiency between 2 locally developed programs and commercial software

TY - JOUR

T1 - First-trimester risk calculation for trisomy 13, 18, and 21: comparison of the screening efficiency between 2 locally developed programs and commercial software

AU - Sørensen, Steen

AU - Momsen, Günther

AU - Sundberg, Karin

AU - Friis-Hansen, Lennart

AU - Jørgensen, Finn Stener

PY - 2011/7

Y1 - 2011/7

N2 - BACKGROUND: Reliable individual risk calculation for trisomy (T) 13, 18, and 21 in first-trimester screening depends on good estimates of the medians for fetal nuchal translucency thickness (NT), free β-subunit of human chorionic gonadotropin (hCGβ), and pregnancyassociated plasma protein-A (PAPP-A) in maternal plasma from unaffected pregnancies. Means and SDs of these parameters in unaffected and affected pregnancies are used in the risk calculation program. Unfortunately, our commercial program for risk calculation (Astraia) did not allow use of local medians. We developed 2 alternative risk calculation programs to assess whether the screening efficacies for T13, T18, and T21 could be improved by using our locally estimated medians. METHODS: We established these estimates from 19 594 women with singleton pregnancies and from 100 pregnant women carrying a fetus affected with trisomy (11 with T13, 23 with T18, and 66 with T21). All measured values were recalculated to a multiple of the median (MoM) and log10 transformed; the mean and SD were calculated for each group. RESULTS: At a given risk cutoff value, we observed a slight improvement in detection rate (DR) for T13, T18, and T21 for a slightly higher false-positive rate (FPR) compared with the commercial program. The lower FPR in the commercial program was caused mainly by an inaccuracy in the PAPP-A median. CONCLUSIONS: Center-specific medians for NT, hCGβ, and PAPP-A should be used in risk calculation programs to ensure high DRs and low FPRs for all 3 trisomies at a given risk cutoff.

AB - BACKGROUND: Reliable individual risk calculation for trisomy (T) 13, 18, and 21 in first-trimester screening depends on good estimates of the medians for fetal nuchal translucency thickness (NT), free β-subunit of human chorionic gonadotropin (hCGβ), and pregnancyassociated plasma protein-A (PAPP-A) in maternal plasma from unaffected pregnancies. Means and SDs of these parameters in unaffected and affected pregnancies are used in the risk calculation program. Unfortunately, our commercial program for risk calculation (Astraia) did not allow use of local medians. We developed 2 alternative risk calculation programs to assess whether the screening efficacies for T13, T18, and T21 could be improved by using our locally estimated medians. METHODS: We established these estimates from 19 594 women with singleton pregnancies and from 100 pregnant women carrying a fetus affected with trisomy (11 with T13, 23 with T18, and 66 with T21). All measured values were recalculated to a multiple of the median (MoM) and log10 transformed; the mean and SD were calculated for each group. RESULTS: At a given risk cutoff value, we observed a slight improvement in detection rate (DR) for T13, T18, and T21 for a slightly higher false-positive rate (FPR) compared with the commercial program. The lower FPR in the commercial program was caused mainly by an inaccuracy in the PAPP-A median. CONCLUSIONS: Center-specific medians for NT, hCGβ, and PAPP-A should be used in risk calculation programs to ensure high DRs and low FPRs for all 3 trisomies at a given risk cutoff.

U2 - 10.1373/clinchem.2010.161299

DO - 10.1373/clinchem.2010.161299

M3 - Journal article

SN - 0009-9147

VL - 57

SP - 1023

EP - 1031

JO - Clinical Chemistry

JF - Clinical Chemistry

IS - 7

ER -