Online Teaching Files

Literature Search and Reviews

Bulletin Board

Reference Charts and Tables

Recommended Links


AFP Testing and the Triple Marker Screen


MS-AFP values (expressed as multiples of the median or MoM) for those fetuses with trisomy 21, spina bifida and those unaffected.


The history of alpha fetoprotein (AFP) testing can be traced back to Great Britain in the 1970s where the incidence of neural tube defects was as high as 5 to 7 per 1000 pregnancies. Neural tube defects (primarily anencephaly, spina bifida or myelomeningocele) are among the most common congenital anomalies occurring in the United States. Anencephaly and spina bifida account for approximately 95% of neural tube defects and encephaloceles make up the remaining 5%.The overall prevalence has been estimated to be 16 per 10, 000 (0.16%) births. The prevalence is higher in the eastern United States and higher in whites and Hispanic women than in other racial or ethnic groups. In California the prevalence is said to be one in 1500 live births.Importantly, the prevalence is higher in children born to families with a history of neural tube defects. In the United States, the risk of recurrence after one child with a neural tube defect is 2% to 3% and 6% after a second abnormal child. Approximately 95% of infants born with neural tube defect are born into families with no previous history of neural tube defects.

The implied rationale for testing or screening programs is to identify those fetuses with neural tube defects and give parents the choice of either continuing with or termination of the pregnancy. As will be discussed below, it became evident soon after its initiation, that other morphologic and chromosomal abnormalities could be detected with such a screening program.

*Note - the author's personal experience with AFP and the Expanded AFP (triple marker) testing program is based upon the California AFP Screening Program. Some of the information is derived from the data and materials from that program. Further information may be obtained from the Genetic Disease Branch, California Department of Health Services.

The identification of the association of elevated levels of AFP and neural tube defects was made over three decades ago. Initially the only mechanism for positive identification of elevated levels of this protein were by sampling amniotic fluid in the second trimester. In the late 1970s the ability to assess abnormal levels in the maternal serum as well, became known. A review of the pathophysiology of alpha fetoprotein is necessary to better understand the advantages and difficulties in such a screening program.

Alpha-fetoprotein (AFP) is a glycoprotein that is produced mainly in the fetal liver, but also in the yolk sac and fetal intestine. The fetal liver produces AFP until 30 weeks of gestation and then stops abruptly. It is related evolutionarily to albumin (similar molecular weight, genes for both proteins are located in chromosome 4, etc). It is found in high concentration in the fetal serum. Circulating time for AFP is 4 days. Serum levels peak at 2 to 3 mg / mL at approximately the 14th week of gestation, after which the level progressively decreases. Normally small quantities of AFP enter the amniotic fluid. While the mechanism of transfer into the amniotic fluid is not fully understood, two likely pathways are: fetal proteinuria and transudation across an immature epithelium. The mechanism of fetal proteinuria is due to the immaturity of the fetal kidneys. As the kidneys mature, the AF concentration gradually decreases. As with the fetal serum, amniotic fluid AFP values peak in the early second trimester and decline as pregnancy progresses. Small quantities of AFP enter the maternal circulation via the placenta. These levels rise progressively from the 7th week to the 32nd week and then decline.

In fetuses with an open neural tube defect (spina bifida) there is absence of the normal integumentary covering at the site of the lesion. This allows abnormally large quantities of AFP to leak into the amniotic fluid and subsequently into the maternal serum. As this phenomenon also occurs in other non-neural tube lesions, other morphologic abnormalities such as abdominal wall defects (AWD) and cystic hygromas likewise demonstrate an elevated level of AFP in the amniotic fluid and maternal serum.

Prior to the observation of the utility of assessing maternal serum values and screening of patients, amniocentesis was the only method for identifying such fetuses with accuracy. The problem however was that method would be impractical for mass screening because of its risks. Likewise, if all the women who had previously been delivered of a child with an open neural tube defect were screened by amniocentesis for elevation of AF-AFP, only 10% of all fetuses with an open neural tube defect would be detected (90% occurring as first time events).

One of the difficult early decisions in initiating a screening program was to decide upon the normal range. Obviously setting a value too low meant too many false positive exams and setting a value too high meant false negative reporting of abnormal cases. Remember that the patient with an abnormal test result will, at the minimum, have increased anxiety, if not proceed to either amniocentesis or termination of pregnancy. At the present time, an maternal serum AFP value is said to be elevated (screen positive) when the value is greater than or equal to 2.5 MoM for a singleton pregnancy and greater than or equal to 4.5 MoM for a pregnancy with two or more fetuses. In most laboratories median concentrations of AFP (and other analytes) are established for the pregnant population for each day between 15 and 20 weeks gestation. The measured value is converted to a "multiple of the median". The median level for each day equals a MoM of 1.00. An AFP result of 1.5 MoM means the patient has one and a half times the median level of AFP, a value of 0.30 MoM means the patient has 30% of the median level of AFP.The values are adjusted for the patients: weight, ethnicity and diabetic status. Heavier pregnant women have lower median values while lighter women have higher values. Black pregnant women have higher median values for AFP. The level of AFP is usually lower in diabetics who are insulin dependent prior to and throughout pregnancy. This does not apply to gestational diabetes. Smoking is also associated with lower MSAFP levels.

In addition to neural tube defects, the causes of elevated maternal serum AFP (MS-AFP) levels include: twins, fetal death, misdated pregnancies, other anatomic abnormalities, placental abnormalities (chorioangiomas, placental lakes, placental edema), fetal growth restriction, preeclampsia, preterm delivery, stillbirth, infection, hypoxia and others. In a study by Reichler et al the authors evaluated 773 patients over a five year period and found that there was a progressive increase in the incidence of anomalies as a direct function of the level of maternal serum AFP, varying from 3.4% at a level of 2.5 MoM to 40.3% at a level of > 7.0 MoM. The increase in the incidence of anomalies is primarily because of the increase in neural tube and ventral wall defects. Divided by types of anomalies, the risks were 4.5% for neural tube defects, 2.6% for ventral wall defects, 2.4% for oligohydramnios and 1.3% for all other anomalies.

The most common reason for false positive results is overlap of the upper end of normal with the lower end of the normal range. In the California program about 1.6% are screen positive for neural tube or abdominal wall defects. In the program participants, the program identifies approximately: 97% of fetuses with anencephaly, 80% of fetuses with open spina bifida and 85% of those fetuses with abdominal wall defects. On the other hand, approximately 9 out of 10 women have screen negative results. While this means that the risk of open neural tube defects and abdominal wall defects are low there still is a chance that the fetus may have these defects.

Low-levels of AFP may be seen in patients who are not pregnant or have a fetal death, have a misdated pregnancy, a hydatidiform mole or trisomic fetus (discussed below) or a normal pregnancy.

It became quickly obvious that ultrasound would play an important role in assisting the diagnosis of abnormal fetuses and helping to correctly assign gestational age. The ultrasound examination was categorized into a level I exam and a level II examination. The purpose of the level I sonogram was to identify misdated pregnancies (18% Calif. program), multiple gestations (10%) and unsuspected fetal death (5%) among women with abnormal MS-AFP levels. It should be noted that while the purpose of the level I exam was not to identify abnormalities, many malformations are detected at the time of this exam. Measurement of the fetal biparietal diameter has been chosen as the method to estimate gestational age. The reasons for this were not only ease of measurement but that fetuses with spina bifida often have smaller than average biparietal diameters. This would make any MS-AFP appear higher than it should be, increasing the sensitivity.

When an abnormal result persists after either repeat of the test or a level I sonogram, the patient may either have an amniocentesis or level II sonogram. This choice varies in different centers. The purpose of the level II examination is to detect a fetal anomaly that might explain the elevation of the MS-AFP. In addition to a general obstetric examination, the neural axis and other anatomic regions known to cause such elevations (abdominal wall, pharynx), should be investigated. Evaluation of the cisterna magna should be performed in all cases. Because open spina bifida lesions are nearly always associated with Chiari II malformations, there is effacement of the cisterna magna in virtually all such cases. The deformity of the frontal calvarium known as the "lemon sign" may also be seen in these fetuses.


Open neural tube defect - spina bifida

Normal cisterna magna. The cisterna magna becomes effaced in the presence of a myelomeningocele

Lemon sign in a fetus with spina bifida

While there is no doubt that the maternal serum AFP screening program has been successful in identifying abnormal fetuses there is still much controversy as to how the program should be administered. There are those who believe that an elevated MS-AFP should be followed by an amniocentesis, those that feel that an elevated MS-AFP should be followed by a level II ultrasound without the need for an amniocentesis and those who believe that sonography could substitute for MS-AFP screening.

Effect of varying the assumed sensitivity of ultrasound on the cost-effectiveness of protocol B(If MS-AFP is greater than or equal to 2.0 MoM targeted ultrasound only) (green line) and the incremental cost of protocol E (if MS-AFP is greater than or equal to 2.5 MoM, amniocentesis for AFP, acetylcholinesterase, and karyotype) versus B (solid blue line) with the cost-effectiveness of the triple panel (dashed line) shown for comparison. The shaded area depicts the sensitivity of ultrasound that was used in this study.

From: Nadel AS, Norton ME, Wilkins-Haug L. Cost-effectiveness of strategies used in the evaluation of pregnancies complicated by elevated maternal serum alpha-fetoprotein levels. Obstet Gynecol 89:660-5, 1997 ©1997 American College of Obstetricians and Gynecologists


Maternal Serum Testing and Chromosomal Abnormalities

The observation of an association between low MS-AFP levels and fetal aneuploidy (independent of age) was first reported in 1984 by Merkatz et al. The interest for the possible relationship between low MS-AFP and trisomy 21 is based upon the fact that only 20% of fetuses with trisomy 21 are found in women undergoing amniocentesis for advanced maternal age or more than 35 years. An additional 20% of fetuses with trisomy 21 are found in women whose MS-AFP level is low after adjustment for age. This relationship has now been confirmed in countless studies and is a part of virtually all maternal serum screening programs. In a review of maternal serum screening in the educational bulletin of the ACOG in September of 1996, the median MS-AFP level in pregnant women carrying fetuses with trisomy 21 was stated to be approximately 0.8 MoM of that of control pregnancies. According to this review, in most laboratories, a patient's result is considered to be positive when the combination of her age and MS-AFP level gives her a risk equivalent to the a priori risk of a 35 year old woman. This is usually a 1:270 midtrimester risk for the occurrence of trisomy 21. In contrast to an absolute cutoff of 2.0 or 2.5 MoM used in neural tube screening, trisomy 21 screening uses a series of age-specific cutoff levels for each MS-AFP level.

In addition to low maternal serum AFP, a number of other analytes have been evaluated as possible markers for chromosomal abnormalities. The two most commonly assessed have been human chorionic gonadotropin (hCG) and unconjugated estriol (µE3). These together with maternal serum AFP constitute the "triple marker screen" or "Expanded AFP screening program". Human chorionic gonadotropin is a hormone synthesized and secreted by the placenta. The levels rise rapidly in early pregnancy and then decline between the 10th and 20th week. Second trimester maternal serum levels may be higher in a pregnancy in which the fetus has trisomy 21. The level of hCG is often lower than usual in a pregnancy with a fetus affected with trisomy 18. Unconjugated estriol is a hormone produced by the fetal adrenal glands, the fetal liver and the placenta. Levels rise throughout normal pregnancy. Maternal serum levels of µE3 may be lower in a pregnancy in which the fetus has trisomy 21 or trisomy 18.

In the California Expanded AFP program a patient is considered screen positive and at an increased risk for carrying a fetus with trisomy 21 when the midtrimester risk is greater tha or equal to 1 in 190. A midtrimester risk of 1 in 190 is equivalent to a term risk of 1 in 250. As was mentioned above, screen positive rates vary by maternal age. Among program participants who were: less than or equal to 34 years of age, 3% were screen positive while those greater than or equal to 35 years of age, 18% were screen positive. The program expects a detection rate of 66% in program participants of all ages for trisomy 21. The detection rate increases as maternal age increases. Low levels of all three analytes are the most common pattern in pregnancies in which a fetus is affected with trisomy 18. A patient is classified as screen positive with an increased risk for carrying a fetus with trisomy 18 when the midtrimester risk is greater than or equal to 1 in 100. As with trisomy 21 screen positive rates vary by maternal age. Among program participants who were: less than or equal to 34 years of age, 0.24% were screen positive, greater than or equal to 35 years of age, 0.35% were screen positive. For all ages the screen positive rate was 0.25%.


Common analyte patterns - Modified from The California Expanded AFP program, Prenatal Provider Handbook


Other Metabolites

The free Beta subunit of hCG has been used as a marker - relative to the whole hCG. The detection rate is similar to the triple screen (58% vs 62%). Dimeric inhibin is another marker being investigated. "It is a placental product which averages 2.06 greater in maternal serum of trisomy 21 fetuses and is independent of AFP, hCG and uE3 levels". (de la Vega) Several investigators have concluded that including it as a fourth metabolite may increase the detection rate over the standard triple screen (Aitken et al, 1996) by as much as 7-12%.


Filly RA, Callen PW, Goldstein RB. Alpha-Fetoprotein screening programs: What every obstetric sonologist should know. In Ultrasonography in Obstetrics and Gynecology, Ed by Peter W. Callen, WB Saunders and Co., Philadelphia, 1994

The Calfiornia Expanded AFP Program, Prenatal Provider Handbook

Maternal Serum Screening,ACOG Educational Bulletin, Number 228, September 1996

Merkatz IR, Nitowsky HM, Macri JN, Johnson WE. An association between low maternal serum alpha-fetoprotein and fetal chromosomal abnormalities. Am J Obstet Gynecol 148:886-891, 1984

Report of UK Collaborative Study on Alpha-Fetoprotein in Relation to Neural Tube Defects. Maternal serum alpha-fetoprotein measurement in antenatal screening for anencephaly and spina bifida in early pregnancy. Lancet 2:1323-1332, 1977

Wenstom KD, Owen J, Davis RO, Brumfield CG Prognostic significance of unexplained elevated amniotic fluid alpha-fetoprotein. Obstet Gynecol 87:213-6, 1996

Main DM, Mennuti MT: Neural tube defects: Issues in prenatal diagnosis and counseling. Obstet Gynecol 67:1, 1986

Milunsky A, Jick SS, Bruel CL et al. Predictive values, relative risks and overall benefits of high and low maternal serum alpha-fetoprotein screening in singleton pregnancies. New epidemiologic data. Am J Obstet Gynecol 161:291, 1989

Crandall BF, Matumota M: Risk associated with an elevated amniotic fluid alpha-fetoprotein level. Am J Med Genetics 39:64, 1991

Shields LE, Uhrich SB, Komarniski CA, Wener MH, Winter TC. Amniotic fluid alpha-fetoprotein determination at the time of genetic amniocentesis: Has it outlived its usefullness?. J Ultrasound Med 15:735-39, 1996

Reichler A, Hume RF, Drugan A, Bardicef M, Isada NB, Johnson MP, Evans MI. Risk of anomalies as a function of level of elevated maternal serum alpha fetoprotein. Am J Obstet Gynecol 171:1052-5, 1994

Nadel AS, Norton ME, Wilkins-Haug L. Cost-effectiveness of strategies used in the evaluation of pregnancies complicated by elevated maternal serum alpha-fetoprotein levels. Obstet Gynecol 89:660-5, 1997 ©1997 American College of Obstetricians and Gynecologists

Nicolaides KH, Campbell S, Gabbe SG, Guidetti R. Ultraound screening for spina bifida: cranial and cerebellar signs. Lancet 2:72-74, 1986

de la Vega A. Understanding maternal serum screening. Lecture. University of Puerto Rico School of Medicine, 2000


This is a clickable imagemap, for hyperlinks, see text below.

Home | Teaching Files | Literature | Bulletin Board


Ultrasound Educational Press     Go Top
Peter W. Callen, M.D.
Professor of Radiology, Obstetrics, Gynecology and Reproductive Science
University of California Medical Center, San Francisco, California