How is maternal Alpha-Fetoprotein (AFP) interpreted?

Interpretation of Maternal Alpha-Fetoprotein (AFP)

Maternal AFP should be reported as multiples of the median (MoM) calculated using laboratory-specific reference data for each gestational week, with values adjusted for maternal weight, race, and insulin-dependent diabetes status before determining patient-specific risks for neural tube defects and chromosomal abnormalities. 1

Essential Clinical Information Required for Interpretation

Before interpreting any maternal AFP result, the following data must be obtained 1:

• Date of birth and last menstrual period to establish accurate gestational age 1
• Ultrasound dating results if available, as this is more accurate than menstrual dating 1
• Maternal weight at time of sample collection, as AFP levels are higher in lighter women and lower in heavier women 1
• Maternal race, particularly Black/African American status, as AFP levels are 10-15% higher in this population 1, 2
• Insulin-dependent diabetes mellitus (IDDM) status, as AFP levels are 10-20% lower in women with IDDM 1
• Date of sample collection 1
• Family history of neural tube defects or chromosomal abnormalities 1

Optimal Timing and Gestational Age Considerations

Screen between 16-18 weeks gestation, though acceptable from 15.0 to 20.9 weeks. 1 Screening performance is significantly decreased at 14 weeks, and results after 20 weeks have limited clinical utility due to restricted management options 1.

Gestational age is the single most critical factor affecting interpretation, as AFP levels rise approximately 16% per gestational week 3. If ultrasound reveals a gestational age discrepancy of 2 or more weeks, the AFP result must be completely reinterpreted using the corrected gestational age 2.

Calculating and Adjusting MoM Values

Step 1: Calculate Raw MoM

The laboratory must establish its own median AFP values for each gestational week (15-20 weeks minimum) using at least 300 samples from unaffected singleton pregnancies 1. The patient's AFP value is divided by the laboratory-specific median for that exact gestational week 1.

Step 2: Apply Adjustment Factors

Multiple correction factors should be applied sequentially, assuming independent effects 1:

• Weight adjustment: Apply published formulas initially, then develop laboratory-specific formulas from in-house data 1
• Race adjustment: Either calculate separate medians for Black/African American women or apply a correction factor (typically 1.10-1.15 multiplier) 1
• IDDM adjustment: Apply correction factor (typically 1.10-1.20 multiplier) to account for lower baseline levels 1

For example, a 200-pound Black woman with IDDM requires all three adjustments applied to her MoM value 1.

Defining Abnormal Results

Elevated AFP (High MoM)

Cutoff levels for open neural tube defect screening are typically 2.0-2.5 MoM in singleton pregnancies and 4.0-5.0 MoM in twin pregnancies. 2 The specific cutoff should depend on patient characteristics (particularly maternal age) and population prevalence of defects 1.

Elevated AFP (both high and low values) may predict serious birth defects or adverse pregnancy outcomes 1.

Low AFP

Use the Down syndrome risk of a 35-year-old woman (1/270 in second trimester) as a reasonable cutoff for defining low values. 1 Low AFP combined with maternal age provides adjusted risk estimates for Down syndrome 1.

Clinical Causes of Elevated Maternal AFP

Fetal Structural Abnormalities

• Open neural tube defects: Detected in 75-90% of cases (95% for anencephaly) 2
• Gastroschisis and ventral wall defects: Associated with very elevated levels 2
• Omphalocele: Normal to elevated levels 2
• Cystic hygroma 2
• Congenital nephrosis: Very elevated levels 2

Other Fetal and Pregnancy Conditions

• Fetal demise 2, 4
• Multiple gestation 1
• Incorrect gestational age (most common cause of apparent elevation) 1

Adverse Pregnancy Outcomes

Elevated AFP without identified structural abnormality is associated with increased risk of low birth weight infants, both from prematurity and intrauterine growth retardation 5. Among infants weighing ≤2.5 kg, 8.5% had AFP >2 MoM; for infants <2 kg, 11.7% had elevated AFP 5.

Management Algorithm for Abnormal Results

For Minimally Elevated AFP (2.0-2.5 MoM)

Obtain a second sample if time permits and the first elevation is minimal. 1 This approach reduces false-positives while maintaining detection rates.

For Significantly Elevated AFP (>2.5 MoM)

Proceed directly to level I ultrasound without waiting for a second sample, especially if pregnancy is relatively advanced or expert sonography is immediately available. 1

After Ultrasound Excludes Dating Error, Twins, and Fetal Demise

Refer promptly to a center with level II ultrasound, amniocentesis capability, and genetic counseling. 1, 2 Both ultrasound levels should be performed by experienced sonographers 1.

The laboratory should telephone all abnormal results directly to the physician and ensure consultative services are available 1.

Critical Pitfalls to Avoid

Gestational age errors are the most common cause of misinterpretation. 2 Always verify dating with ultrasound before proceeding with invasive testing.

Inadequate clinical information (missing weight, race, or diabetes status) leads to incorrect MoM calculations and risk assessment 2. The laboratory should refuse to interpret results without complete data 1.

Using reference data from other laboratories rather than establishing laboratory-specific medians significantly degrades screening performance 1. Each laboratory must develop its own reference ranges for its population and update them periodically 1.

Samples drawn before 15 weeks may give inaccurate results and should be interpreted with extreme caution 2.

Laboratory Reporting Requirements

The laboratory report must include 1:

• Quantitative AFP level in ng/mL or IU/mL 1
• MoM value calculated from laboratory-specific medians 1
• Patient-specific risk for open neural tube defects 1
• Age-specific and adjusted risk for Down syndrome based on AFP and maternal age 1
• Statement about clinical relevance to neural tube defects and chromosomal abnormalities 1

Laboratories must participate in external quality control programs and document satisfactory performance 1.