The triple-marker test in predicting fetal aneuploidy: a compromise between sensitivity and specificity

Area of Science:

• Prenatal diagnostics
• Biochemical screening
• Genetics

Background:

• Down syndrome (trisomy 21) is a common chromosomal abnormality.
• Prenatal screening aims to identify pregnancies at increased risk for Down syndrome.
• Triple-marker screening (alpha-fetoprotein, human chorionic gonadotropin, unconjugated estriol) is a widely used method.

Purpose of the Study:

• To evaluate the role of unconjugated estriol in Down syndrome detection.
• To assess the impact of maternal age, cut-off choice, and population specificity on screening performance.
• To optimize the balance between sensitivity and specificity in the triple-marker test.

Main Methods:

• Analysis of serum markers (alpha-fetoprotein, human chorionic gonadotropin, unconjugated estriol) in 2833 pregnant women.
• Calculation of trisomy 21 risk using default and population-specific medians.
• Testing different risk cut-offs (1:300, 1:200, 1:100) and evaluating marker contributions via logistic regression.

Main Results:

• Unconjugated estriol exclusion reduced detection rates by 33% and improved specificity by 4%.
• A cut-off of 1:300 yielded 87.5% sensitivity and 63.3% specificity; a 1:100 cut-off yielded 66.7% sensitivity and 79.5% specificity.
• Unconjugated estriol was the only marker significantly classifying patients into normal and trisomic groups (trisomy 21 and 18).

Conclusions:

• Unconjugated estriol is crucial for Down syndrome detection, despite a minor decrease in specificity.
• A 1:300 risk cut-off presents an unfavorable sensitivity-specificity balance.
• A 1:100 cut-off is recommended for higher-risk pregnancies, potentially guiding amniocentesis decisions.