Prenatal diagnosis for paediatricians

Abstract

In Ontario, approximately 140,000 women deliver newborn infants each year. Of these women, 60,000 to 70,000 have multiple marker screening, 10,000 undergo amniocentesis or chorion villus sampling and virtually all have at least one prenatal ultrasound. Multiple marker screening is not used in every province and territory; however, amniocentesis and prenatal ultrasound are used throughout Canada. Most paediatric patients will have been exposed to some form of prenatal diagnosis. If an abnormality is found prenatally, parents may have concerns to discuss with the paediatrician after the child is born. Likewise, if a child with a problem is born following a normal pregnancy, the parents will want to know why the problem was missed prenatally. Paediatricians should be aware of prenatal tests that have been performed to understand better their patients and their families.

Prenatal diagnosis is pervasive. In Ontario, approximately 140,000 women deliver each year. Of these, 60,000 to 70,000 have multiple marker screening, 10,000 have amniocentesis or chorion villus sampling and virtually all have at least one prenatal ultrasound. While multiple marker screening is not used in every province and territory, amniocentesis and prenatal ultrasound are used throughout Canada. Consequently, most paediatric patients will have been exposed to some form of prenatal diagnosis. When all goes well, both pre- and postnatally, this is of little importance to the paediatrician. However, when an abnormality is found prenatally, it is likely that parents will have concerns to discuss with the paediatrician postnatally. Likewise, if a child with a problem is born following a normal pregnancy, the parents will want to know why the problem was missed prenatally. Therefore, while it may seem somewhat odd for an article on prenatal diagnosis to appear in a paediatric journal, it is important for paediatricians to be aware of testing that has happened during gestation to better understand patients and their families.

DIAGNOSIS VERSUS SCREENING

Prenatal diagnosis is generally divided into diagnostic and screening modalities, with the latter being the most commonly used. Some of these modalities, such as maternal serum screening, are highly specific for disorders such as Down syndrome and spina bifida, while others such as ultrasound are much less specific, looking for a whole range of anomalies. These modalities will be discussed.

The purpose and value of screening and diagnostic tests in prenatal diagnosis are frequently confused both by patients and health care providers. In general, screening is designed to identify people at increased risk within a normal population. Screening may be as simple as taking a family history or asking a woman her age, or it may be more complex (eg, using multiple markers to screen for Down syndrome). The information garnered from screening is used in directing prenatal testing. Diagnostic testing, as the name would suggest, is used to rule in or out a diagnosis in someone who has screened ‘positive’.

SCREENING TESTS

Family history

The first step in prenatal screening should be a three generation family history taken from the pregnant woman or woman considering a pregnancy. This would include information about previous children, stillbirths, abortions and miscarriages. The woman should be asked specifically about birth abnormalities, chromosomal anomalies, mental retardation or developmental delay, hereditary diseases and consanguinity.

Teratogen history

Preferably before she gets pregnant, a woman should be asked about her medications, smoking, alcohol and street drug use as well as her immunizations. She should also be questioned about her workplace to assess environmental exposures. An attempt can then be made to eliminate any potential teratogens before conception.

Frequently, when a fetus has been exposed to a teratogen, a risk can be assigned and testing such as ultrasound can be done. A normal test result may reassure the parents but often the risk cannot be eliminated. This is often a source of ongoing anxiety. It is important to keep in mind that whenever a child is found to have a congenital or genetic problem, one of the most common questions from parents will be about possible gestational exposures.

Screening based on ethnicity

The ethnicity of both the woman and her partner should be asked. In Canada, women are screened for the hemoglobinopathies – sickle cell anemia and thalassemia in non-Northern European populations, and for Tay-Sachs disease in the Ashkenazi Jewish population (1). In some centres, screening for Canavan disease and familial dysautonomia is also offered to Jewish patients. Other jurisdictions may do more extensive ethnic screening (2).

Screening parameters

Traditionally in prenatal screening, particularly with multiple markers, slightly different screening parameters have been used. Rather than sensitivity and specificity, the terms detection rate (DR) and false positive rate (FPR) are used. The DR is the number of affected pregnancies with positive results and is equal to the sensitivity. The FPR is the number of unaffected pregnancies with positive results or 1 – specificity. The FPR and DR will vary with the risk cut-off that is chosen. If a low risk cut-off is chosen, the DR will be increased, but so will the FPR because the net is being thrown more widely. Similarly, if a high risk cut-off is chosen, then fewer women will be termed positive unnecessarily, but more cases will be missed. The best screen will have the maximum DR for the minimum FPR.

The other difference in prenatal screening is the method of reporting marker levels. The most common way of doing this is in multiples of the median. For each marker, a median curve against gestation is developed and then each individual is reported as a multiple of the median for her gestation. The idea behind this is that results would be equivalent from one laboratory to another and independent of assay variations among laboratories (3).

Screening for open neural tube defects

In the presence of an open neural tube defect (ONTD), a fetal liver protein known as alpha-fetoprotein (AFP) leaks out into the amniotic fluid and passes into the maternal circulation, raising the maternal serum AFP (MSAFP) levels (3). The measurement of MSAFP is gestation-dependent (4). Depending on the risk cut-off, MSAFP screening detects about 80% of ONTDs, with a 2% FPR (5).

Elevated MSAFP levels are most commonly due to multiple gestation pregnancies (6), fetal demise (7) and underestimated gestation, but can also be seen with other fetal anomalies including abdominal wall defects, skin disorders and congenital nephrosis (8–10). Unexplained elevations of MSAFP are associated with an increased risk of intrauterine growth restriction, oligohydramnios, fetal demise and maternal pre-eclampsia (11–14).

If the MSAFP is positive for an ONTD, generally a detailed ultrasound is offered although amniocentesis would be an option. If amniocentesis were to be done, the amniotic fluid would be sent for AFP testing as well as acetylcholinesterase testing. Acetylcholinesterase is an enzyme that is elevated in the presence of an ONTD. Although it is somewhat dependent on the laboratory, ultrasound will generally detect approximately 90% to 95% of cases of ONTD (15). Amniocentesis will detect about 95% of ONTDs (16). MSAFP and amniocentesis will not detect closed neural tube defects.

Folic acid:

While folic acid therapy is not part of prenatal diagnosis, it is used to prevent birth defects, particularly ONTDs. It is recommended that all women considering a pregnancy take 0.4 to 1.0 mg folic acid/day for at least two months before conception and continue until the second missed period (17,18).

Screening for Down syndrome

Until the mid-1980s, the only screen for Down syndrome was maternal age. Since that time, a number of different screens have been developed (Table 1). Some of these screens are not used in Canada, but may be introduced in the future.

TABLE 1.

Comparison of screening modalities

Screening modality	False positive rate (%)	Detection rate (%)
Maternal age	11.9	39.8
Maternal serum alpha-fetoprotein	12.3	49.8
Maternal serum screening	7.4	70.6
First trimester screening	5	80
Integrated prenatal screening	1.5	83

Maternal age:

Between the mid-1960s and the mid-1980s, the method of screening for Down syndrome was to ask a woman her age. If she was 35 years or older, she had a positive screen and was offered amniocentesis. If maternal age screening is applied to the current population in Ontario, 40% of cases of Down syndrome would be detected with an FPR of 12%. Only one in 150 screen positive women would have a baby with Down syndrome.

MSAFP:

In 1983, Merkatz et al (19) reported an association between a low MSAFP and fetal chromosomal anomalies. Subsequently, MSAFP combined with maternal age was used as a biochemical marker for Down syndrome as well as ONTDs. It was a slight improvement on maternal age alone (Table 1) (20).

Maternal serum screening

In 1988, Wald et al (21) suggested combining two new markers, estriol and beta-human chorionic gonadotrophin with MSAFP in the detection of Down syndrome. This is called triple marker or maternal serum screening (MSS). MSS can also be used in the detection of trisomy 18 (22,23). A provincial MSS program using triple markers was started in Ontario on July 1, 1993. Results from the first six years of this program have shown a 70.6% DR for Down syndrome with a 7.4% FPR with a risk cut-off of one in 385 at term. If the result is positive for Down syndrome or trisomy 18, amniocentesis is offered. Overall, women with a positive MSS result have a one in 44 chance of having a baby with Down syndrome. While this means that the majority of women who have a positive MSS result will have a normal baby, there is evidence to suggest that women are still concerned about the baby’s health even when the karyotype is normal (24).

First trimester screening

While MSS has been an improvement on previous screening modalities for Down syndrome, there have been a number of valid criticisms. One major concern for both women and health care providers has been the FPR of 7.4% (25,26). Another concern has been the relatively late identification of fetal anomalies. It has been suggested that it would be far less distressing for women if chromosomal anomalies could be detected in the first trimester (27,28). For this reason, both first trimester ultrasound and biochemical markers have been developed.

Nuchal translucency:

Both children and fetuses with Down syndrome have been noted to have thickening of the skin at the back of the neck. Nicolaides et al (29) used this information to screen for Down syndrome. In a study of 96,000 pregnancies they estimated that nuchal translucency (NT) alone would detect 82% of cases of Down syndrome with an 8.3% FPR (30). The DR is probably closer to 70% (31), but this still makes NT the most powerful marker for Down syndrome. There is also evidence to suggest that NT is useful in the detection of other chromosomal anomalies, birth defects such as congenital heart disease and certain genetic syndromes (32). Unfortunately, even using all the available prenatal tests, it is impossible to detect all the anomalies potentially associated with an enlarged NT and, thus, very difficult to fully reassure parents that the result is a false positive.

First trimester biochemical markers:

A number of biochemical markers have been investigated in the first trimester. To date, the most useful markers are pregnancy-associated plasma protein A and free beta-human chorionic gonadotrophin, which is biochemically related to marker beta-human chorionic gonadotrophin tested in the second trimester. While neither of these markers is effective alone, when combined with NT, the DR can be improved to 80% for a 5% FPR (33).

Integrated prenatal screening

While first trimester screening has improved both DR and FPR, many women who undergo NT screening still request MSS in the second trimester. This has created a great deal of confusion because, until recently, it has not been possible to combine these results. If the two screens were discordant, neither the woman nor her health care provider knew which to believe. Wald et al (34) suggested a method of integrating first and second trimester results that not only circumvented the problem of two separate results, but also increased the DR of Down syndrome while significantly decreasing the FPR (Table 1). The major drawback of this method is that to keep the FPR low, women (and their providers) have to agree to wait until the analysis of the second trimester markers to receive a result. Wald et al (34) predicted that integrated screening would have an 85% DR with a 1.5% FPR. Screen-positive women would have a one in 10 chance of having a baby with Down syndrome.

Ultrasound soft signs

A number of ultrasound findings in the second trimester have been associated with Down syndrome and other chromosomal anomalies. These include choroid plexus cysts, mild ventriculomegaly, increased nuchal thickening, echogenic cardiac focus, renal pyelectasis, echogenic bowel, increased fetal iliac angle, shortened femurs and humeri, ear length and fifth finger clinodactyly (35–41). While all of these findings may increase the risk for a chromosomal anomaly, the actual risk for each finding is not entirely clear. In addition, it is not known if these risks are independent of multiple marker risks. Therefore, it is not advisable to use these findings to numerically modify the risk obtained through previous screening. The only possible exception would be the relationship between choroid plexus cysts and trisomy 18, which has been reasonably well quantified (42).

Certain soft signs such as mild ventriculomegaly, renal pyelectasis and echogenic bowel have implications beyond the chromosomal risk.

Mild ventriculomegaly:

When a fetus has been found to have isolated mild ventriculomegaly (width of the atria of the lateral ventricles is 10 to 15 mm), the mother will be offered viral testing and amniocentesis for karyotyping. While there is very poor follow-up of these cases, even when the tests are normal, there is probably a risk for developmental delay (3% to 36%) (43,44).

Renal pyelectasis:

Renal pyelectasis is associated with a relatively low risk for Down syndrome, and one of the major concerns with this marker is when it persists into the newborn period. Pyelectasis should always be followed up with a renal ultrasound in the first week of life after feeds have been established. If there is evidence of reflux, then a urology consult should be considered (45,46).

Echogenic bowel:

As well as being associated with chromosomal anomalies, echogenic bowel can be associated with viral infections and cystic fibrosis (CF) (47,48). Generally, both will be tested during pregnancy, but in molecular testing for CF, it is only possible to test for the most common mutations in a particular ethnic group. This means that, unless there are known mutations in the family, the laboratory cannot state that a fetus does not carry a CF gene. Therefore, in the relatively rare scenario that a child tested in utero is showing signs of CF, the paediatrician should not feel completely reassured by the prenatal test results.

DIAGNOSTIC TESTS

Chorion villus sampling

Chorion villus sampling (CVS) is usually performed at approximately 11 weeks’ gestation and involves transcervical or, less commonly, transabdominal sampling of the early placenta (chorionic villi). CVS tends to be reserved for women at higher risk or for those who require fetal molecular testing, but this depends on the centre. There is a one in 100 risk of miscarriage (49) and a small risk of terminal transverse limb defects in the fetus that is probably most significant when CVS is performed before 11 weeks’ gestation (50). CVS is a focused test, meaning that a normal result rules out a chromosomal anomaly and/or a particular genetic syndrome, but it does not guarantee the birth of a normal baby.

Amniocentesis

Amniocentesis involves transabdominal sampling of amniotic fluid usually after 14 weeks’ gestation. A Canadian randomized trial of early amniocentesis (11 to 13 weeks’ gestation) showed an increased risk for clubfoot among exposed fetuses (51,52). There is currently a trial based in the United States looking at the risks associated with 13- to 14-week amniocenteses. Amniocentesis is associated with a one in 200 risk of miscarriage (53,54). Amniotic fluid is routinely sent for chromosome and AFP testing; however, a variety of tests including metabolic and molecular studies can be done on either the fluid or the amniocytes. Like CVS, amniocentesis is focused on a specific question and a normal result does not guarantee the birth of a normal baby.

Ultrasound

In most cases, ultrasound is used as a screen; however, sometimes when a clear fetal anomaly, such as a neural tube defect, omphalocele, hydrocephalus, etc, is found, it may also be diagnostic. A detailed ultrasound at 18 weeks’ gestation has been recommended by the Society of Obstetricians and Gynecologists of Canada for all pregnant women (55).

Fetal echocardiography

In many cases a fetal cardiac screen done using ultrasound will be adequate and echocardiography will be unnecessary. A fetal echocardiogram should be considered if a previous child or one of the parents was born with a heart defect; if the fetus has a disorder known to be associated with a cardiac anomaly such as Down syndrome, 22q13 deletion; or if a cardiac abnormality is detected on ultrasound screening (56).

CONCLUSION

While certain prenatal information such as teratogen exposure, a clear anomaly on ultrasound or an abnormal karyotype may provide direct information, the paediatrician must also be aware of the indirect effects of prenatal screening and testing. Many women who have had a positive MSS or a soft sign found on ultrasound with a subsequent normal karyotype feel that there still must be some sinister reason for the positive result. For the paediatrician to have some insight into this attitude, it is important that he or she has access to the prenatal records and understands what has happened before the child became his or her patient.