- Top of page
In The Netherlands, transabdominal two-dimensional (2D) ultrasound screening during the second trimester of pregnancy was made universally accessible by legislation and health insurance coverage in 2007. Evaluation of the upper lip for possible cleft lip and palate is an optional element in this screening (NVOG Dutch Society of Obstetrics & Gynaecology protocol1). With an incidence of cleft lip with or without cleft palate (CL ± P) and cleft palate only (CP) of 1.8 per 1000 births2, this translates to approximately 320 cases of orofacial cleft per year in The Netherlands3. As 90–95% of Dutch pregnant women undergo ultrasound screening, cleft palate teams are commonly confronted with prenatally diagnosed clefts.
Classification of orofacial clefts diagnosed by ultrasound is generally carried out according to the system of Nyberg et al.4. However, in this classification no distinction is made between the alveolus and the hard palate, which, in the presence of CL, can be important in the detection of an additional CP. Proper classification is essential because different types of cleft may be variably associated with additional anomalies and chromosomal disorders4–7, while isolated clefts are associated with low mortality and morbidity rates and are primarily a functional and esthetic problem8.
In a recently published systematic review it was concluded that transabdominal 2D ultrasound screening in a low-risk population has a relatively low rate of detection of orofacial clefts8, with detection rates of 23–58% in the most recent large prospective studies9, 10. Moreover, in most studies reporting on the accuracy of ultrasound screening, inconsistent methodological quality is observed. In this prospective cohort study, the sensitivity and specificity of ultrasound in detecting orofacial clefts during prenatal screening in a low-risk and a high-risk population is presented, as well as the percentage of associated anomalies per type of orofacial cleft.
- Top of page
This prospective follow-up study investigated the detection of orofacial clefts by ultrasound screening in The Netherlands during the period of January 2007 to December 2008. For the purpose of screening, The Netherlands is subdivided into eight regions, in which trained midwives, sonographers and obstetricians perform routine 2D ultrasound screening in several centers. Every region has one main referral center to which pregnant women are referred if an abnormality is suspected. Upon referral, certified obstetricians perform further 2D and three-dimensional (3D) ultrasound examinations, and the patient is counseled by members of the cleft palate team11. The standard for advanced 2D ultrasound examination is set by the Dutch Society for Gynaecology and Obstetrics (NVOG), the documentation of which is available electronically1. Ultrasound screening routinely includes detailed examination for orofacial clefts.
The study cohort comprised all screened pregnant women in the geographically well-defined screening region of Utrecht. The study population was divided into a low-risk population and a high-risk population. This separation was made to ensure that all pregnant women in each group had a comparable a priori risk. Pregnant women who underwent routine screening formed the low-risk population, and the high-risk population consisted of pregnant women with a positive family history of orofacial cleft or other risk factors for fetal abnormalities. The high-risk population and fetuses suspected of orofacial cleft were seen in a tertiary referral center (the Wilhelmina Children's Hospital). When an associated anomaly was detected, karyotyping by amniocentesis and consultation with a clinical geneticist was recommended. When parents opted for termination of pregnancy (TOP), mostly because of the presence of associated anomalies, an autopsy was recommended to verify the prenatal diagnosis. When parents opted to continue a pregnancy, they met a multidisciplinary team, which included an obstetrician and two members of the cleft palate team: a plastic surgeon and a medical psychologist. In the first week after delivery all newborns were examined by a plastic surgeon and a pediatrician or a clinical geneticist from the cleft palate team to verify prenatally detected orofacial clefts. Newborns in the region of Utrecht who had an orofacial cleft that was undiagnosed prenatally were referred to the cleft palate team at the Wilhelmina Children's Hospital and evaluated at the cleft clinic within 2 weeks after delivery.
Data on all ultrasound examinations were retrieved from the screening centers where data were stored electronically (MOSOS version 7–9, Copyright @ 1988–2009; Bureau Medische Automatisering, Houten, The Netherlands). Postnatal data were retrieved from the clinical records of the cleft palate team. Statistical analysis was performed on anonymized data using SPSS statistical software (Version 14; SPSS Inc., Chicago, IL, USA). The significance level was set at P < 0.05.
- Top of page
During the 2-year study period a total of 35 924 low-risk pregnant women and 2836 high-risk pregnant women underwent ultrasound screening in one of the 25 non-profit centers belonging to the Wilhelmina's Children's Hospital referral network. The resulting study population of 38 760 women represented 93% of the total pregnant population of 41 482.
Sixty-two fetuses (liveborn, stillborn or TOP before 24 weeks) with confirmed orofacial clefts were identified in the study population. This resulted in a prevalence of 1.6 per 1000 fetuses (including isolated and associated cases). For liveborn infants the prevalence in this group was 1.42. The distribution in the total population was 29% (18/62) cleft lip (CL), 40% (25/62) cleft lip with cleft palate (CLP), 27% (17/62) CP, one median cleft and one atypical cleft (4%) (Table 1). Male to female ratios for each cleft type are shown in Table 2. The median gestational age at detection of an orofacial cleft was 21 (range, 18–23) weeks of gestation. An additional five cases of facial cleft were identified postnatally by the cleft palate team in patients whose mothers had not undergone prenatal ultrasound screening.
Table 1. Overview of distribution of all facial clefts in total population showing performance of ultrasound screening in each risk group
| || || ||Low-risk group||High-risk group|
|Cleft type||n||Prevalence per 1000*||Sensitivity (% (n))||False positive (n)||Sensitivity (% (n))||False positive (n)|
|CL||18||0.46||81 (13/16)||1||100 (2/2)||0|
|CLP||25||0.64||91 (20/22)‡||0||100 (3/3)‡||1|
|CP||17||0.44||0 (0/14)||1||0 (0/3)||0|
|Atypical cleft||1||0.03||100 (1/1)||0||—||0|
|Total||62||1.60||65 (35/54)||2||62.5 (5/8)||1|
Table 2. Male to female ratio per cleft type
|Cleft type||n||Male/female ratio|
Twelve of the 62 fetuses (19%) with orofacial cleft in the study population had a positive family history; in nine cases there was a first-degree relative (mother, father or sibling) with a CL ± P and in three cases a second-degree relative (aunt, uncle or cousin). Seven pregnant women with a positive family history, and so who should have been referred directly to a specialized center and screened as high risk, were screened in the low-risk clinics (four had first-degree and three had second-degree relatives with facial clefting).
A total of 54 fetuses with clefts were in the low-risk population (Table 1). There were 14 cases of CP, none of which was diagnosed prenatally. CL was detected prenatally in 13/16 cases, giving a sensitivity of 81% (95% CI, 56–94%). Two false-positive cases were detected in the low-risk group. One case was a fetus with multiple congenital anomalies, and karyotyping showed trisomy 13. The parents decided on TOP and the facial cleft was not confirmed by autopsy. In the second case a CP was suspected after screening by 2D ultrasound. A normal lip and palate were observed on specialized ultrasound scans in the Wilhelmina's Children's Hospital and this was confirmed postnatally. CLP was detected prenatally in 20/22 cases, giving a sensitivity of 91% (95% CI, 69–98%). One case of midline cleft was correctly detected during screening, along with additional anomalies (hypertelorism, agenesis of the corpus callosum or encephalocele). One fetus had a Tessier IV cleft with microphthalmia, and the parents decided on TOP.
A total of eight fetuses with clefts were in the high-risk population. They all underwent at least two 2D and 3D second-trimester scans in the Wilhelmina Children's Hospital. There were three cases of CP, none of which was diagnosed prenatally (Table 1). CL was detected prenatally in 2/2 cases and CLP was detected in 3/3 cases, giving a sensitivity of 100% (95% CI, 38–100%) in both groups. There was one false-positive case, concerning a fetus with multiple congenital anomalies and trisomy 18, in which the parents decided on TOP and the CLP was not confirmed during autopsy.
Of 62 fetuses with a facial cleft, 24 (39%) had associated anomalies (Table 3). CL had a low percentage (17% (3/18)) of associated anomalies, whereas bilateral CLP and CP had much higher incidences of associated anomalies (57% and 53%, respectively).
Table 3. Associated and isolated orofacial clefts per cleft type
| ||Isolated cleft||Associated cleft|
|Cleft type||n (%)||Prenatally detected (n)||TOP (n)||IUD (n)||Postpartum death (n)||n (%)||Prenatally detected (n)||TOP (n)||IUD (n)||Postpartum death (n)|
|CL||15 (83)||13||—||—||—||3 (17)||2||1||—||—|
|CLP (uni)||12 (67)||11||—||—||—||6 (33)||6||2||1||—|
|CLP (bi)||3 (43)||3||1||—||—||4 (57)||3||—||2|| |
|CP||8 (47)||0||—||—||—||9 (53)||0||—||—||—|
|Total||38/62 (61)|| ||1||—||—||24/62 (39)|| ||4||3||—|
There were eight cases with chromosomal abnormalities, in six of which prenatal karyotyping had been performed for suspicious ultrasound findings; in the remaining two cases the diagnosis was made postnatally. The association of chromosomal and syndromal anomalies in each group of facial cleft type is shown in Table 4. Trisomy 18 was the most common chromosomal defect observed. Almost all syndrome diagnoses were in the group of CP, with an incidence of nine out of 17 (53%) in this group. One case of Goldenhar syndrome was diagnosed in combination with a Tessier IV cleft. In the CP group there were three fetuses with Pierre–Robin sequence. Figure 1 shows the incidence of chromosomal anomalies according to the type of cleft and the presence or absence of additional anomalies on prenatal ultrasound screening, illustrating the risk of missing a diagnosis of a chromosomal anomaly in these cases. In the group of bilateral CLP, trisomy 18 was diagnosed postnatally and no associated anomalies were observed on ultrasound.
Figure 1. Incidence of chromosomal anomalies according to type of cleft and presence or absence of additional anomalies on prenatal ultrasound screening, illustrating the risk of missing a diagnosis of a chromosomal anomaly in these cases. *In one case with trisomy 18 and bilateral cleft lip with cleft palate (CLP), the anomalies and CLP were missed on ultrasound at 20 weeks, although they were detected on ultrasound at > 24 weeks of gestation. CL, cleft lip; MC, median cleft.
Download figure to PowerPoint
Table 4. Summary of anomalies, syndromes and chromosomal defects per cleft type
|Cleft type||Anomaly||Syndromes/sequence||Chromosomal defects||Prenatally detected|
|CL||Umbilical hernia|| || ||No|
| ||Club foot|| || ||No|
| || || ||Trisomy 18||Yes|
|CLP (uni)||Hypospadias|| || ||No|
| ||Hirschsprung's disease|| || ||No|
| ||VSD|| ||Trisomy 21||Yes (both anomalies)|
| || || ||Trisomy 18 (× 3)||Yes (× 3)|
|CLP (bi)||Hypertelorism, ACC|| || ||Yes|
| || || ||Trisomy 13||Yes|
| || || ||Trisomy 9p||Yes|
| || || ||Trisomy 18||No|
|CP|| ||Pierre–Robin (× 3)|| ||No (× 2)*, Yes (× 1)|
| || ||Stickler syndrome (× 2)|| ||No (× 2)|
| || ||Kabuki syndrome (× 2)|| ||No (× 2)†|
| || ||van der Woude syndrome|| ||No|
| || ||Beckwith–Wiedemann syndrome|| ||No‡|
|MC||Hypertelorism, ACC, encephalocele|| || ||Yes|
|Atypical cleft||Tessier IV cleft, microphthalmia||Goldenhar syndrome|| ||Yes|
Fifty-four infants (87%) were born alive (Table 3). TOP was performed in five cases. In all but one terminated fetus there were severe associated anomalies present, the exception being a case of isolated CLP. Three cases of intrauterine death were registered in fetuses with trisomies 13 (n = 1) and 18 (n = 2). All surviving infants were still alive at the time of writing and under the care of the cleft palate team.
- Top of page
Through introduction of routine prenatal screening, cleft palate teams are more frequently confronted with prenatally diagnosed clefts. Increased awareness of facial clefting and associated anomalies might clarify pregnancy options. When a diagnosis is made, parents have the opportunity to prepare themselves emotionally and practically12 by meeting a multidisciplinary team consisting of an obstetrician, a psychologist and a plastic surgeon. This allows the parents to discuss the surgical options as well as the demands of a growing child with a cleft lip and palate.
This prospective study of ultrasound screening demonstrates a sensitivity of 88% for CL ± P; however, the overall sensitivity for orofacial clefts was lower—65% in the low-risk group and 62.5% in the high-risk group—because no cases of isolated CP were diagnosed prenatally. Pregnant women for whom there was suspicion of a fetal facial cleft were referred to a referral center for 2D and 3D ultrasound scans where obstetricians detected additional cleft palates in three of six cases (50%) in which there was initially a cleft only lip identified by primary ultrasound. These rates are high in comparison with former studies that reported variable orofacial cleft detection rates of 23–59.6% on 2D ultrasound and 66–100% on 3D ultrasound8, 13. In our study, sources of bias were minimized by determining the diagnostic accuracy of screening tests8. A large study population in a well-defined geographic area was prospectively enrolled. Also, high-risk and low-risk pregnant women were studied separately. The relative risk of familial recurrence was recently reported to be 32 (95% CI, 24.6–40.3) for any CL in first-degree relatives and 56 (95% CI, 37.2–84.8) for cleft palate alone14, 15. Postnatal follow-up was at least 1 year, long enough to detect the majority of associated anomalies. Ultrasound evaluation of the study population was carried out according to a standardized protocol. To perform routine prenatal screening, minimal quality criteria were set for sonographers (with a minimum of 150–200 examinations annually)11. Consequently, obstetricians performing ultrasound examinations in a high-risk setting have more experience and the opportunity to use 3D ultrasound techniques, which resulted in the detection of additional cleft palates in three of six cases (50%). This is in concordance with Crane et al., who reported significantly higher detection rates in tertiary centers vs. non-tertiary centers16.
Three cases of false-positive diagnoses were identified; in two fetuses with multiple congenital anomalies and trisomy 13 and 18, respectively, an ultrasound diagnosis of CLP was not confirmed by autopsy. In the third case, CP was suspected on screening by 2D ultrasound, but subsequent referral revealed no CP on specialized ultrasound. The specificity of prenatal ultrasound will be high because of the large group of true negatives and it is therefore not informative to calculate diagnostic accuracy. The impact of a falsely diagnosed anomaly can be severe.
It is important to differentiate correctly between different types of facial clefts because they are variously associated with additional anomalies4, 6, 7. Previous prenatal studies have quoted rates of from 0%6 to 20%4 for CL, 24%7 to 52%6 for unilateral CLP and 48%7 to 79%6 for bilateral clefting, which are in concordance with our results. Cases of CL were associated with almost no chromosomal or syndromal defects (6%), CLP (especially bilateral) had 28% chromosomal or syndromal defects and CP had a high percentage of syndromal defects (53%) but no chromosomal defects. The percentage of associated anomalies is higher in cases with prenatally detected orofacial clefts than in those only identified postnatally, because of spontaneous fetal death and medical intervention. Eighty-eight percent of the chromosomal defects were detected prenatally, with one case of trisomy 18 missed. Additional anomalies in cases with chromosomal defects are usually more evident and detection of one should prompt more detailed screening, also leading to increased detection of orofacial clefts. As different types of orofacial cleft are variously associated with additional anomalies, not every pregnant woman with a diagnosis of a cleft should receive further prenatal invasive testing owing to the minor risk of miscarriage. All patients with detected clefts should be referred to a clinical geneticist. There is a minor chance (3%) of chromosomal abnormality when no associated anomalies are diagnosed, in agreement with a recent study by Gillham et al.13. Therefore, we advise no further analysis in cases with suspected isolated CL. In isolated CLP, one should be cautious with invasive diagnostics. When associated anomalies are detected before 24 weeks, one should discuss the option of karyotyping and multiplex ligation-dependent probe amplification (MLPA) for 22q11 deletions and fluorescence in-situ hybridization (FISH) (a molecular cytogenetic technique used to detect chromosomal microdeletions that may be missed on routine chromosome analysis)17. As median clefting is never found in isolation, cytogenetic and molecular diagnostic analysis should be offered.
The consequences of routine ultrasound at 20 weeks of gestation are the subject of ongoing political discussion in The Netherlands. Recently, the numbers of terminations performed in the hospital setting in the second trimester (12–24 weeks) were published, revealing an increase of 140 (7.4% of all TOPs) to 270 (11.5% of all TOPs) in 2005–2007, after the introduction of standard ultrasound scans18. However the trend towards more hospital terminations (i.e. on medical grounds) started in 2003 and is attributed to the increase in ultrasound screening, even before the current legislation was introduced18. Moreover, termination for isolated CL ± P remains uncommon in The Netherlands. In the region of Utrecht, TOP for isolated orofacial clefting occurred only once in this period. Apparently, parental consideration is different in terms of minor and major deformities, especially when a deformity is readily amenable to reconstruction19.
In conclusion, this study cohort of a non-selected population of 38 760 pregnant women included 62 fetuses with facial clefts. Thirty-nine per cent of these had associated anomalies. There was a high sensitivity of detecting orofacial clefts by ultrasound screening. The high rate of detection of an orofacial cleft is probably a result of the excellent training and experience of the sonographers, referral to specialized centers when a cleft is suspected, routine visualization of the fetal face and advances in ultrasound techniques. Different types of orofacial cleft relate differently to associated anomalies and thus to ultimate outcome. For this reason, pregnancies in which there is suspicion of a fetal cleft with associated anomalies should be routinely referred for further genetic analysis.