Routine Fetal Anomaly Scan

Introduction

Since the introduction of ultrasound scanning into obstetric practice in UK in the 1980s, there has been a considerable change to the role of this test in antenatal care, which has been driven largely by developments in technology. In most units, it is now part of routine screening in pregnancy and its main purpose is considered to be the detection of fetal structural anomalies. However, until recently, there was little uniformity in how and when the scan should be carried out, which structures should be examined and which abnormalities might be identified.

This chapter aims to address the questions of why, when and how the anomaly scan should be conducted and to review its effectiveness in the detection of fetal anomalies. We will discuss the following questions:

What is the purpose of the second-trimester scan?
When is the optimal time for the second-trimester anomaly scanning?
What is the detection rate for specific major structural malformations from routine second-trimester screening?
Which structures should be examined and what are the minimum standard views during a routine second-trimester screening examination?
What is the role of audit?

What is the Purpose of the Second-Trimester Anomaly Scan?

The role of the second-trimester scan has evolved, especially in countries where a universal 11–13 weeks scan has been introduced. The National Institute for Health and Care Excellence (NICE) UK recommends that all pregnant women be offered two scans in pregnancy, the first at 11–13 weeks and a second one around 20 weeks. The objectives of the first scan are to establish gestational age, detect multiple pregnancies with determination of chorionicity and to measure fetal nuchal translucency (NT) thickness as part of combined screening for trisomy 21. As such, the main focus of the second-trimester scan has become the identification of fetal anomalies, although it still plays a more general role in the assessment of fetal biometry, placental site and amniotic fluid volume.

It should be emphasized that the routine second-trimester scan is a screening test. Screening programmes are intended to be applied to the general population, or a selected subgroup. The aim of a screening programme is early identification of a preventable or treatable condition where there is a useful intervention which should result in a reduction in mortality and/or morbidity. Perinatal mortality has fallen in developed countries since 1950, with the relative contribution of congenital malformations increasing substantially. The prevalence of fetal defects is around 2% with congenital malformations accounting for 20–30% of neonatal deaths. Whilst the introduction of routine second-trimester scanning might be expected to improve the detection of fetal anomalies, translation of this into measurable reduction in perinatal mortality or morbidity is more complex.

The RADIUS study (Routine Antenatal Diagnostic Imaging with Ultrasound) published in 1993 in the USA concluded that routine ultrasound, compared with selective use of sonography, did not improve perinatal outcome. This was a highly controversial study with a poor sensitivity in detection of fetal anomalies (16.6% before 24 weeks’ gestation). The Helsinki Ultrasound trial was the first to report a clear reduction in perinatal mortality by 47% in singleton and 58% in twin pregnancies. Bricker et al., in a systematic review of routine second-trimester ultrasound, concluded that the routine anomaly scan can reduce the perinatal mortality if firstly, the detection of fetal malformations is an important objective, secondly, if this is matched with a high level of diagnostic expertise and finally, if termination of pregnancy for fetal abnormality is widely accepted in the population screened.

In the 2010 NICE publication on Antenatal Care, the purpose of the routine second-trimester scan was defined as being the identification of fetal anomalies to allow:

Reproductive choice.
Parents to prepare for any treatment/disability/palliative care or consider termination of pregnancy.
Management of the pregnancy/delivery in a specialist centre.
Intrauterine therapy.

Therefore, screening should target conditions that are either incompatible with life or associated with severe morbidity, those that may benefit from antenatal intervention or that require early intervention following delivery. The National Screening Committee Fetal Anomaly Screening Programme (FASP) UK have developed recommendations for which key fetal structures should be examined leading to the identification of any one of 11 specific conditions. These conditions have been chosen either because they are fatal or associated with important morbidity or require immediate postnatal support. Only conditions with reported detection rates (DR) of at least 50% from a commissioned review of the scientific literature were included ( Table 3-1 ). They concluded that whilst other conditions may also be detected during screening, there was insufficient published data to guide the standards of DR to be set.

TABLE 3-1

The Conditions that Should Be Screened for as a Minimum and Respective Detection Rate (NHS FASP)

Conditions	Detection Rate (%)
Anencephaly	98
Open spina bifida	90
Cleft lip	75
Diaphragmatic hernia	60
Gastroschisis	98
Exomphalos	80
Serious cardiac abnormalities	50
Bilateral renal agenesis	84
Lethal skeletal dysplasia	60
Edwards syndrome (trisomy 18)	95
Patau syndrome (trisomy 13)	95

Women's Views

Women's understanding of the purpose of the scan is variable – in the authors' experience, it is very often considered a chance to see the baby and confirm normality rather than a screening test to look for abnormalities. A review of women's views on pregnancy ultrasound included studies from 18 countries and concluded that ultrasound is very attractive to women and families, but women often lack information about the purposes for which an ultrasound scan is being done as well as the technical limitations of the procedure. NICE recommends that women should be ‘informed of the limitations of routine ultrasound screening and that detection rates vary by the type of fetal anomaly, the woman’s body mass index and the position of the unborn baby at the time of the scan’. Women should receive comprehensible information before the scan that will allow them to make an informed decision whether to have the test and if a woman chooses to decline, her choice should be respected.

When is the Optimal Time for the Second-Trimester Anomaly Scan?

The optimal timing of the routine second-trimester anomaly scan must be a balance between the earliest gestation at which the maximum number of main fetal structures can be adequately assessed, and the latest gestation at which an acceptable range of options can be offered to parents, if an anomaly is detected. There is no international consensus regarding the specific time for the routine second-trimester anomaly scan, in the time interval between 18 and 22 weeks. Both NICE and FASP recommend it is carried out between 18 weeks 0 days and 20 weeks 6 days, with a single further scan at 23 weeks to complete the screening examination if the image quality of the first examination is compromised by increased body mass index, uterine fibroids, abdominal scarring and/or sub-optimal fetal position. The Fetal Medicine Foundation suggests that in settings where first trimester screening is implemented at 11–13 weeks, and dating, chorionicity and major malformations have already been addressed, the 20–22 weeks interval performs better for anomaly screening.

It has been shown that in an unselected pregnant population, second-trimester ultrasound screening is easier to perform and less likely to require additional scan at 20–22 weeks than at 18 weeks. The incidence of low-lying placenta was also higher at 18 weeks than at 20 and 22 weeks in this study ( Table 3-2 ). As such, it is likely that units performing the anomaly scan at 18–20 weeks rather than at 20–22 increase their workload due to inability to complete the anomaly screening in one visit and follow-ups for low lying placenta.

TABLE 3-2

Percentage of Low-Lying Placenta and Anomaly Scan Completed by Gestation

Gestation	18–18 ⁺⁶ weeks	20–20 ⁺⁶ weeks	22–22 ⁺⁶ weeks
% (n) Anomaly scan completed	76.3% (306)	90% (371)	88.8% (393)
% (n) Low-lying placenta *	20.6% (83)	15.1% (62)	13% (52)

* Lower edge less than 10 mm from the internal os.

Individual units should decide the optimal time for their scan, taking into account the main purpose of the scan and their local facilities. If the anomaly scan is the first scan in the pregnancy, 18–20 weeks may have advantages in dating and identifying chorionicity in multiple pregnancies, whereas in centres where first-trimester screening is implemented, the 20–22 weeks interval may be preferable. Centres need to balance the time needed for counselling and further investigations with the ethical and legal restrictions in case a malformation is identified and termination of the pregnancy is considered.

What is the Detection Rate for Specific Major Structural Malformations from Routine Second-Trimester Screening?

Sensitivity and Specificity

In a systematic review of studies in the 1990s on the clinical effectiveness of the routine second-trimester scan, Bricker et al. reported that the overall prevalence of fetal abnormalities was 2% and about 45% of these were detected by a routine scan, with a wide variation of 15–85.3%.

There are many factors that may account for this wide variation in sensitivity, including inconsistency in screening protocols, gestation at examination, experience of the operator and sophistication of equipment. Also the postnatal ascertainment of abnormalities can have a major influence on the detection rate of routine screening, such that performance of ultrasound can be overestimated, where ascertainment is poor, particularly in anomalies that are not externally visible, such as cardiac septal defects and renal abnormalities. For the assessment of fetal and neonatal deaths, postmortem examinations are the gold standard but these are not performed routinely everywhere. Finally the population sample may also result in dramatically different detection rates. Screening pregnant women at high risk is likely to be much more effective since the chance of finding an anomaly is higher, examiner concentration is better and operators are generally better trained. Most birth defects occur, however, in an unpredictable way in couples with no particular risk and therefore the approach discussed here is to offer ultrasound to look for structural defects in all pregnant women.

For the purpose of this chapter we have confined our discussions on sensitivity and specificity to studies where large population samples were used, and specifically address detection rates for screening before 24 weeks’ gestation when data are available ( Table 3-3 ).

TABLE 3-3

Second-Trimester Studies Reporting the Detection of Fetal Anomalies Before 24 Weeks' Gestation Using Routine Ultrasound

Study	Period	GA at Scanning	Number of Fetuses	Number of Abnormal Fetuses	Number Detected	Prevalence Abnormalities (%)	Sensitivity (%)	Specificity (%)
Rosendahl & Kivinen	1980–1988	<24	9012	93	37	1.0	39.4	99.9
Chitty et al.	1988–1989	18–20	8785	125	93	1.5	74.4	99.9
Levi et al.	1984–1989	16–22	15654	381	154	2.4	40.4	100
Shirley et al.	1989–1990	18–20	6412	84	51	1.4	60.7	99.9
Luck et al.	1988–1991	18–20	8844	166	140	1.9	84.3	99.9
Roberts et al.	1988–1989	<24	12909	192	85	1.7	44.3	?
Crane et al.	1987–1991	15–22	7685	187	31	2.5	16.6	?
Levi et al.	1990–1992	16–22	9392	235	120	2.5	40.5	99.9
Papp et al.	1988–1990	18–20	51675	496	317	2.3	63.9	100.0
Anderson et al.	1991–1993	16–20	7880	144	84	1.8	58.3	?
Carrera et al.	1970–1991	<22	33192	1006	598	3.0	59.4	?
Geerts et al.	1991–1992	18–24	457	10	5	2.8	50.0	?
Boyd et al.	1991–1996	<24	33376	725	328	2.2	45.2	99.5
Smith & Hau	1989–1994	<24	246481	818	434	1.1	53.0	?
Eurenius et al.	1990–1992	15–22	8324	145	32	1.7	22.1	99.8
Grandjean et al.	1990–1993	<24	170800	3685	1754	2.2	44.0	?
Stefos et al.	1990–1996	18–22	7236	162	130	2.2	80.3	99.9
Wong et al.	1993–1998	16–24	12169	169	123	1.4	72.8	?
Tabor et al.	1997–1998	18–20	7963	100	66	1.3	64.1	?
Garne et al.	1995–1999	<24	1198519	4567	1900	0.4	43.5	?
Nackling & Backe	1989–1999	18	18181	267	103	1.5	38.6	99.9
Richmond & Atkins	1985–2000	<24	573471	12648	6544	2.2	51.7	?
Fadda et al.	1981–2004	<24	42256	1050	578	2.5	55.1	99.88

GA, gestational age.

In the late 1990s there were two studies based on large population samples: the EUROFETUS and the EUROSCAN. Both were European studies with the aim of evaluating the prenatal detection rate of fetal anomalies by routine ultrasonographic examination in unselected populations. Both studies covered a 3-year period (1990–1993 for EUROFETUS and 1996–1999 for EUROSCAN) with more than 100,000 women screened. There are however some differences that are important to highlight : EUROFETUS was a prospective study including 61 obstetric units in 14 countries in Europe (institution-based) where there was at least one systematic examination of the fetus preferably between 18 and 22 weeks. In the EUROSCAN study the data were collected retrospectively from 20 registries of congenital malformations from 12 European countries (geographical area-based) with different policies, from no routine fetal scanning (Netherlands and Denmark) to one (UK) or more than one routine fetal scans in pregnancy (France, Germany, Italy, Spain). Also the reported frequency of malformed fetuses in the EUROSCAN study is 1.15%, significantly less than 2.2% in EUROFETUS. The EUROFETUS study reported an overall detection rate of malformed fetuses of 61.4% with 55% of the fetuses with major defects detected before 24 weeks. The EUROSCAN detection rate was not given as a whole but only for specific classes of anomalies. The study reported wide variations in the detection rates between different registries and these were lower in countries with no routine ultrasound scanning policy. However, there were no differences in detection between countries performing one scan in the second trimester and those scheduling two or more scans.

EUROCAT (European Concerted Action on Congenital Anomalies and Twins) is a European network of population-based registries for the epidemiologic surveillance of congenital anomalies. EUROCAT began in 1979 and the most recent data available spans the period 2006–2010. It now includes 43 registries in 23 countries with more than 1.7 million births surveyed per year, representing 29% of the European birth population. EUROCAT has two main objectives: to provide essential epidemiologic information on congenital anomalies in Europe and to facilitate the early warning of new teratogenic exposures.

In 2005, Garne et al. published the EUROCAT data from 1995 to 1999 focusing on 11 severe congenital malformations. The overall prenatal detection rate was 64% (range, 25–88% across regions) with 68% of cases (range 36–88%) diagnosed before 24 weeks (calculated detection rate of 43.5% from the data available). As with the previous two studies there was significant variation between different countries and the probability of identifying a malformation increased with an increasing number of malformations. Moreover, there were also large differences in overall detection rates according to the type of anomaly (96% for anencephaly and 33% for transposition of the great vessels).

Specificity of routine second-trimester screening is not reported in many of the published studies. Bricker et al., in their systematic review reported false-positive rates between 0% and 0.55%. The EUROFETUS study reported an overall false-positive rate of 9.9% and distinguishes these from false alarms when the diagnostic error was rectified on subsequent ultrasonographic examination at a later stage of the pregnancy. Some of the more common conditions that were associated with a false-positive diagnosis included microcephaly, anomalies of the aorta, hydronephrosis, oesophageal atresia and talipes.

Topographic Detection of Fetal Anomalies

We have previously identified that comparisons of overall detection rates for screening by routine second-trimester scan are unhelpful as it is clear that detection rates vary significantly depending on the type of anomaly being screened for. Therefore, it makes more sense to compare published data for specific abnormalities, where available. In this section the detection of fetal abnormalities will be discussed in relation to the individual anatomical compartments.

It should be appreciated that the studies which inform this section were mostly carried out in the late 1990s. Since this time, bodies such as RCOG, NHS FASP and more recently ISUOG have published guidelines on the minimum standards for the ultrasound examination of the fetus and there has been increasing focus on training and the development of increasingly sophisticated ultrasound technology. These factors, as well as the introduction of the routine first-trimester screening for chromosomal defects in many countries, should result in an overall improvement in detection rates in current times, compared with previously published data.

Skull, Brain and Spine

The most frequently diagnosed abnormalities of the fetal central nervous system are hydrocephalus, spina bifida (including meningocoele or meningomyelocoele), and anencephaly, followed by encephalocoele, holoprosencephaly and Dandy–Walker malformation.

Ventriculomegaly occurs in 1% of pregnancies and refers to enlargement of the cerebral ventricles, whereas hydrocephalus is a pathological enlargement of the ventricles and occurs in only 1 in 2000 pregnancies. Hydrocephalus is often associated with other conditions, namely spina bifida. In such cases the fetal head circumference is often normal before 24 weeks. The overall rates of detection of fetal hydrocephalus before 24 weeks' gestation using mid-trimester ultrasound examination are reported to vary from 33–100%. This reflects the diverse aetiology of the condition and the fact that the sonographic features may not be apparent until the late second or early third trimester.

Neural tube defects (NTD) occur in about 5 per 1000 pregnancies and include anencephaly, encephalocoele and spina bifida. The vast majority of NTD are spina bifida and anencephaly with only 5% being encephalocoeles. Detection rates are generally high in this group of malformations. Anencephaly is reliably diagnosed in the second trimester as it has a classic ‘frog eyes’ appearance. The fetal skull ends just above the orbits and this is coupled with secondary brain erosion. Although reported detection rates vary from 80–100%, the vast majority of centres have a rate approaching 100%. Encephalocoeles are cranial defects with herniated fluid-filled or brain-filled cysts, most commonly occurring in the occipital area. The detection rates for this condition are difficult to interpret due to the low frequency of the defect, but are reported to be 60–100%, the lower figure probably being the most accurate estimate, as it comes from the larger studies. Open spina bifida is suspected when there is disruption of the normal skin covering the spine, splaying of the vertebrae in the coronal view, the presence of a myelocoele or myelomeningocoele, or associated cranial signs (‘lemon’ and ‘banana’ signs). Ventriculomegaly is also found in about 70% of cases of open spina bifida in the mid-trimester. The reported detection rates for open spina bifida before 24 weeks vary from 66–93%.

Holoprosencephaly consists of a spectrum of disorders that result from incomplete cleavage of the forebrain. It is a rare defect, occurring in about 1 in 10,000 births. Reported detection rates therefore suffer from the same limitations as others with low frequencies but are reported as being 69–100%.

Dandy–Walker complex refers to a spectrum of defects of the cerebellar vermis, cystic dilatation of the fourth ventricle and enlargement of the cisterna magna. Dandy–Walker malformation is rare and occurs in about 1 in 30,000 livebirths. Reported detection rates are difficult to interpret due to study design and large variations in the reported frequency of the condition, but are reported to be 41–93%.

Face

Facial clefting occurs in about 1 in 700 births and is the most common facial defect detected at the routine mid-trimester fetal anomaly scan. Clefting may be unilateral, bilateral or midline and can involve a wide spectrum of facial anatomy including the uvula, soft palate, hard palate, alveolar ridge and upper lip. In approximately half of all cases both the lip and palate (CLP) are involved, and in 25% of cases each, the defect is isolated to either the lip (CL) or the palate (CP). The condition is unilateral in the majority (75%) of cases and usually isolated (80%).

The reported detection rate for isolated CL or CLP at a routine mid-trimester scan (between 18–24 weeks) in an unselected population, varies widely from 18–88%. Studies reporting the highest detection rates suggest that this may be attributable to the routine inclusion of specific views (coronal view of the lips, transverse view of the alveolar ridge ± profile view). In the EUROSCAN study in which 709,027 pregnancies were examined in 12 European countries between 1996–1998, the overall detection rate for CL/CLP was 21% (161 of 751). However in cases of isolated CL/CLP the detection rate dropped to 17% compared with 44% for cases associated with chromosomal defects, multiple malformations or genetic syndromes. Variation in detection rate between individual centres was wide (0–75%) and may be explained by differences in operator expertise, local protocols, gestational age at assessment or the sophistication of ultrasound equipment. There were no reported cases of a false-positive diagnosis in any of these studies other than that of Stefos et al., in which there was a single case.

Additional facial anomalies that may be identified at the mid-trimester scan include orbital and ophthalmic defects (hypo/hypertelorism, micropthalmia/anopthalmia), micrognathia, frontal bossing and nasal bone hypoplasia, but there are no published studies of the detection rate at routine anomaly screening.

Thorax

The most common anomaly of the chest is congenital diaphragmatic hernia (CDH), which occurs in approximately 1 in 4000 pregnancies. Left-sided lesions are more easily identified due to the contrast in echogenicity of the small bowel and the lung and/or by the presence of an echolucent stomach ‘bubble’ within the chest. In right-sided CDH, the liver can appear similar in echogenicity to the lung. Approximately half of all cases of CDH are isolated and these may be less frequently detected than those associated with other anomalies or chromosomal defects. The reported overall detection rate for CDH at a mid-trimester scan (before 24 weeks) varies from 36–71%. Those studies with the lowest rates reported improving detection over time. The figures are also higher if later gestational ages are included. In the EUROFETUS study, there were 187 cases of CDH with a detection rate of 59% (110 of 187) but only half of these were diagnosed before 24 weeks' gestation. This study also reported a significant difference in detection rate between isolated cases (51%) and those with associated malformations (72%).

Other defects which may be detectable include pleural effusions, cystic lesions of the lungs, the most common being congenital cystic adenomatoid malformation (CCAM) (1 in 4000), and less commonly, bronchogenic cysts, pulmonary sequestration and bronchial atresia. However, it has been observed that the ability to correctly diagnose cystic lesions of the lungs by prenatal ultrasound is limited and that the sonographic features may not be predictive of the postnatal histological classification into CCAM or sequestration. Furthermore, published studies are concerned with postnatal outcome and there is insufficient data to assess the detection rate of cystic anomalies at routine mid-trimester sonography.

Heart

Congenital heart defects occur in 5–10 per 1000 live births. The aim of routine screening should be to identify serious cardiac defects as evidence suggests that antenatal diagnosis of major congenital heart disease improves outcomes, particularly in relation to secondary cerebral injury. Prenatal diagnosis allows for appropriate and timely intervention in a hospital where specialist cardiothoracic surgery is available. Routine examination of the four-chamber view (see Figure 3-11 ) should exclude abnormalities that result in discrepancy in the size of the ventricles and/or impaired function of atrio-ventricular valves. Normal offsetting of the AV valves should exclude AVSD and an intact interventricular septum excludes a significant VSD. Extending the examination to include the outflow tracts has been shown to improve detection of major cardiac defects. Conotruncal defects such as tetralogy of Fallot, transposition of the great arteries, truncus arteriosus and double outlet right ventricle are identified from these additional views.

In the UK, data from the National Institute for Cardiovascular Outcomes Research, reported a detection rate for serious cardiac defects (those requiring surgery or therapeutic catheterization in the first year of life), which has risen from 23% in 2004 to 35% in 2011. There is a large variation in detection rates worldwide and data from the EUROSCAN group reported rates of 11–48%. Unsurprisingly, these variations seemed to be explained by differences in the availability of routine screening. In Western Europe, detection rates were highest in countries with three routine scans and lowest in those with none. Furthermore, detection was lower with isolated cardiac malformations and higher in lesions specifically affecting the size of the ventricles. It has been long established that there is an association between increased nuchal translucency and cardiac defects. A recent meta-analysis suggested that detection rates might be improved to 50% by using first-trimester nuchal translucency screening to identify a high-risk group. However, current NICE guidelines do not recommend this policy.

Abdomen

The most commonly identified abdominal defects are gastroschisis and exomphalos, each occurring in 1 in 4000 pregnancies. In gastroschisis, there is a small abdominal wall defect, lateral and usually to the right of an intact umbilical cord insertion, through which there is evisceration of the gut into the amniotic cavity. In exomphalos, there is herniation of the abdominal contents, which can include liver, bowel and spleen, covered by peritoneum, into the base of the umbilical cord. The reported detection rates before 24 weeks’ gestation for both are high, approaching 80–100% in most large series.

Gastrointestinal obstruction (GIO) may result from atresia, stenosis, agenesis or fistula of any part of the intestine including oesophagus, duodenum, and small or large intestine. Collectively GIO occurs in about 1 in 200 births, but is infrequently detected at the mid-trimester scan, as many features do not manifest until the third trimester, and some are not diagnosable before birth. In the EUROSCAN study, the overall detection rate before 24 weeks was 14% (48 of 349) but this varied with the subtype of GIO. Detection rates before 24 weeks were 7% for oesophageal, 19% for duodenal and 12% and 20% for small and large bowel respectively. In all cases, detection rates were higher later in gestation. Oesophageal atresia may be suspected if there is inability to visualize the fetal stomach, however there is often an associated tracheo-oesophageal fistula, which allows the stomach to fill via the trachea, making the diagnosis elusive. Polyhydramnios, which develops well after the time of the routine scan, may be the only clue to the underlying pathology. Duodenal atresia may be suspected by the presence of a ‘double bubble’ representing a dilated duodenum and stomach but the associated polyhydramanios usually occurs after 26 weeks.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here