What does magnetic resonance imaging add to the prenatal ultrasound diagnosis of facial clefts?

Authors


Abstract

Objective

Ultrasound is the modality of choice for prenatal detection of cleft lip and palate. Because its accuracy in detecting facial clefts, especially isolated clefts of the secondary palate, can be limited, magnetic resonance imaging (MRI) is used as an additional method for assessing the fetus. The aim of this study was to investigate the role of fetal MRI in the prenatal diagnosis of facial clefts.

Methods

Thirty-four pregnant women with a mean gestational age of 26 (range, 19–34) weeks underwent in utero MRI, after ultrasound examination had identified either a facial cleft (n = 29) or another suspected malformation (micrognathia (n = 1), cardiac defect (n = 1), brain anomaly (n = 2) or diaphragmatic hernia (n = 1)). The facial cleft was classified postnatally and the diagnoses were compared with the previous ultrasound findings.

Results

There were 11 (32.4%) cases with cleft of the primary palate alone, 20 (58.8%) clefts of the primary and secondary palate and three (8.8%) isolated clefts of the secondary palate. In all cases the primary and secondary palate were visualized successfully with MRI. Ultrasound imaging could not detect five (14.7%) facial clefts and misclassified 15 (44.1%) facial clefts. The MRI classification correlated with the postnatal/postmortem diagnosis.

Conclusions

In our hands MRI allows detailed prenatal evaluation of the primary and secondary palate. By demonstrating involvement of the palate, MRI provides better detection and classification of facial clefts than does ultrasound alone. Copyright © 2010 ISUOG. Published by John Wiley & Sons, Ltd.

Introduction

Orofacial clefts are congenital malformations caused by abnormal facial development during gestation, and occur in about 1/700 live births1. The severity of a cleft is related to its extension, varying from cleft lip alone to complete cleft of the lip, alveolus and palate2. The overall prognosis of facial clefts depends on whether there are associated severe additional malformations, which occur in 21% of cases of cleft lip and palate and in 8% of cases of cleft lip alone3.

Ultrasound, as the primary imaging method, provides real-time examination, is easily accessible and is inexpensive4. However, the prenatal sonographic detection of cleft lip and palate can be difficult because of shadowing from the surrounding bony structures5. The sensitivity of two-dimensional transabdominal ultrasound examination for facial clefts was reported in 2000 to be around 27% for cleft lip and palate and 7% for isolated cleft palate in the routine obstetric screening population, depending on the experience of the examiner and the type of malformation4. Due to technical improvements, especially of three-dimensional ultrasound technology, evaluation of the secondary palate is now possible in selected cases6, 7 and orofacial pathologies may be detected even during first-trimester screening8.

Magnetic resonance imaging (MRI) is a valuable complement to sonography5, adding useful information about the maxillofacial anatomy and allowing precise evaluation of the primary and secondary palate9. The purpose of this study was to investigate the ability of fetal MRI of the maxillofacial region in visualizing the location and extent of malformations as well as the accuracy of this method in detecting associated anomalies.

Methods

In this retrospective study, which was approved by our institutional review board, we searched our hospital database for all facial clefts diagnosed with fetal MRI in our department over an 8-year period (2001–2008). Thirty-four such cases were identified and included in the study. They had been referred for fetal MRI due to findings on routine ultrasonography of suspected facial cleft (n = 29), micrognathia (n = 1), cardiac defect (n = 1), brain anomaly (n = 2) or diaphragmatic hernia (n = 1). Before the MRI examination, all patients were informed about the procedure and its safety10.

MRI was performed on a 1.5-Tesla superconducting unit using a five-element, phased-array surface coil with women positioned in a supine position, or, in cases of polyhydramnios or multiple pregnancy, in a left lateral decubitus position. T2-weighted, ultrafast spin-echo sequences, echoplanar sequences and steady-state free-precession sequences with 2–4-mm slice thickness were obtained. We used axial views at the level of the maxilla, coronal views at the level of the fetal palate, including the lip region, and sagittal views. On the basis of embryological development, clefts were grouped into the following categories: isolated cleft lip; cleft lip and alveolus; cleft lip and palate; and isolated cleft palate (Figure 1). Clefts were further subdivided into categories of extension: unilateral or bilateral. All women were offered the option of fetal karyotyping.

Figure 1.

View of the primary and secondary palate, demonstrating a cleft of the soft palate alone (a), a cleft of the hard palate alone (isolated cleft palate) (b), a bilateral cleft lip and palate (c), a unilateral cleft lip (d), a unilateral cleft lip and alveolus (e) and a unilateral cleft lip and palate (f).

Postnatal diagnosis was determined by the operating surgeon or by the pathologist in cases of intrauterine fetal death or termination of pregnancy, and compared with the intrauterine MRI findings. In addition, the prevalence and prenatal detection rate of facial clefts was determined. All cleft types were recorded either as an isolated finding or as associated with other structural anomalies.

Results

Among the 34 patients, there were 11 (32.4%) clefts of the primary palate (cleft lip/alveolus) alone, 20 (58.8%) clefts of the primary and secondary palate (cleft lip and palate), and three (8.8%) isolated clefts of the secondary palate (cleft palate) identified by MRI. The mean gestational age at the time of ultrasound examination was 24 (range, 14–33) weeks and at MRI it was 26 (range, 19–34) weeks. Patient characteristics, types of cleft, associated anomalies and pregnancy outcomes are given in Table 1.

Table 1. Comparison of prenatal sonographic diagnosis with prenatal magnetic resonance imaging (MRI) findings and postnatal diagnosis in 34 cases of fetal cleft lip (CL) or palate (CP)
nSonographic diagnosis before referral for MRIMRI diagnosisSonography vs. MRIAdditional structural anomalies (on US* and/or MRI) and chromosomal aberrations
  • *

    Detected on ultrasound (US).

  • Detected on magnetic resonance imaging (MRI).

6CLCLUS findings confirmed (n = 6)Mild Turner syndrome (n = 1), ventriculomegaly* and club feet (n = 1), cardiac defect* and duodenal stenosis (n = 1)
3CLCL, cleft alveolusMisclassification on US (n = 3)Hydrocephalus* (n = 1)
8CLCL/CPMisclassification on US (n = 8)Ventriculomegaly and hypertelorism (n = 1), polycystic kidneys (n = 1)
1CLIsolated CPMisclassification on US (n = 1)Holoprosencephaly* (n = 1)
1CL/CPCL, cleft alveolusMisclassification on US (n = 1) 
8CL/CPCL/CPUS findings confirmed (n = 8)Ventriculomegaly and microphthalmia and hypotelorism (n = 1)
2CL/CPIsolated CPMisclassification on US (n = 2)Ventriculomegaly and cardiac defect* (n = 1), holoprosencephaly* (n = 1)
4Microcephalus (n = 1), hydrocephalus (n = 1), cardiac defect (n = 1), diaphragmatic hernia (n = 1)CL/CPMissed on US (n = 4)Microcephalus* and microphthalmia (n = 1), holoprosencephaly (n = 1), diaphragmatic hernia* and cardiac defect* (n = 1), cardiac defect (n = 1), deletion chromosome 15 (n = 1), duplication chromosome 14q (n = 1)
1MicrognathiaCLMissed on US (n = 1) 

The detection rate for facial clefts was 100% (n = 34) using MRI as all the cases were selected from the MRI database. Of these, prenatal ultrasound examination identified the presence of cleft lip/palate in 29 (85.3%) cases; in the remaining five (14.7%) cases, patients were referred because of other structural anomalies seen on ultrasound (Table 1) and the clefts were diagnosed only by MRI. Of the 29 cases with facial cleft suspected initially on ultrasound, the interpretation of the referral ultrasound findings correctly identified the extent of palate involvement in only 14 (41.2%) cases, with 15 (44.1%) facial clefts being misclassified. However, this fact did not modify clinical management in any case.

Of the 11 (32.4%) infants which had only a cleft primary palate, this was detected on prenatal ultrasound in six. On MRI examination, clefts of the primary palate were best visualized in the axial and coronal views of the fetal face. The axial planes were used to evaluate the soft tissue of the upper lip, the six anterior tooth buds and the underlying alveolar ridge. A cleft lip with cleft alveolus was suspected in the presence of a disrupted alveolar ridge (Figure 2). Additional findings on MRI, i.e. those not found on ultrasound examination, included detection of club feet (n = 1) and duodenal stenosis (n = 1).

Figure 2.

T2-weighted axial magnetic resonance images showing a unilateral cleft at 22 gestational weeks (a), 25 weeks (b), 34 weeks (c) and 35 weeks (d). The arrows mark the disrupted alveolar ridge with missing or abnormally arranged tooth buds.

There were 20 (58.8%) fetuses with a cleft of the primary and secondary palate; these were best visualized on MRI in the axial and coronal planes (Figure 3). Unilateral clefts of the primary and secondary palate were identified in the presence of a disrupted irregular contour of the alveolar ridge or in the presence of the characteristic deviation of the nasal septum shown in the axial images. A bilateral defect was suspected in cases of soft tissue protrusion of the premaxillary segment, which was best visualized on axial and sagittal views. Missing or abnormally implanted anterior tooth buds in the alveolar ridge confirmed the presence of a cleft of the primary and secondary palate; ultrasound correctly identified eight of these. Fetal MRI correctly classified the extent and the laterality in all 20 clefts of the primary and secondary palate. In seven of these 20 fetuses, additional malformations involving the central nervous system (n = 4), the heart (n = 2) and the fetal kidneys (n = 1) were revealed by MRI, while these findings were identified by ultrasound in only three fetuses.

Figure 3.

T2-weighted axial magnetic resonance images of fetal unilateral cleft lip and palate at 22 weeks (a) and 25 weeks (b). The arrows mark the defect in the soft tissue of the upper lip.

In three (8.8%) cases, we found an isolated cleft of the secondary palate. None of these was detected by ultrasound, the patients having been referred for MRI because of structural anomalies detected during ultrasound examination (holoprosencephaly (n = 2), ventriculomegaly and cardiac defect (n = 1)). In contrast to sonography, MRI allows direct visualization of the secondary palate, which was possible in all three of these cases (Figure 4). The secondary palate was best seen when amniotic fluid filled the fetal mouth, allowing precise visualization of the oropharynx and tongue. The secondary palate was best visualized on MRI in axial and sagittal sections that showed it as a continuous straight hypointense line (Figures 5 and 6).

Figure 4.

T2-weighted sagittal magnetic resonance image of an isolated cleft of the secondary palate at 24 weeks' gestation. The arrow marks the abnormally elevated tongue movement into the nasal cavity.

Figure 5.

T2-weighted sagittal magnetic resonance image of a normal palate at 25 weeks' gestation. The arrow marks the hypointense line of the soft palate.

Figure 6.

(a,b) T2-weighted sagittal magnetic resonance images of a complete cleft of the primary and secondary palate in two fetuses at 23 (a) and 22 (b) weeks. The arrows mark the absence of the hypointense layer of the palate. (c,d) T2-weighted coronal magnetic resonance images of two fetuses at 23 (c) and 22 (d) weeks with a bilateral cleft of the primary and secondary palate. The arrows mark the bilateral cleft.

Nineteen (55.9%) fetuses were karyotyped prenatally, and three (15.8%) of these fetuses had chromosomal aberrations. Pregnancy was terminated in four (11.8%) cases, including the three fetuses with chromosomal aberrations and one fetus with severe additional malformations.

The perinatal mortality rate was 16.7% (5/30). Three (8.8%) fetuses died in utero and two (5.9%) infants died during the neonatal period because of severe additional anomalies. In all 34 cases, the MRI diagnosis was confirmed either by pediatric examination or autopsy. The location of the cleft and its degree of extension into the secondary palate was correctly detected in all fetuses at MRI examination, as verified in the neonatal period (Figure 7). There were no false-positive diagnoses. All surviving infants underwent a normal procedure for cleft repair.

Figure 7.

Unilateral cleft of the lip and primary palate at 33 weeks' gestation shown on sagittal (a) and axial (b) T2-weighted magnetic resonance imaging and postnatally (c). In (a), the intact secondary palate is visualized in the sagittal plane as a straight hypointense line (arrow). The axial image (b) shows the irregular contour of the alveolar ridge (arrow). (c) The unilateral cleft of the primary palate was diagnosed on day 1 after delivery.

Discussion

The prospect of delivering a child with a facial cleft is associated with considerable stress for the parents11. Cleft lip and palate are embryologically distinct processes, varying from a submucous malformation to a complete cleft of the primary and secondary palate, with differing prognoses12. The importance of assessing the degree of extension of a facial cleft and the presence of any associated severe, and possibly lethal, anomalies is obvious and prenatal detection offers the potential for optimal planning of neonatal management13.

Ultrasound is the imaging modality of choice for evaluation of the fetus, and antenatal detection of cleft lip and palate using two-dimensional ultrasound has been described14. Clefts of the primary palate may be detected routinely, whereas clefts of the secondary palate can remain undetected15, especially if they appear without any additional malformations4. The Eurofetus study of routine ultrasonography in pregnancy showed a detection rate for cleft lip of 25%, for cleft lip and palate of 22% and for isolated cleft palate of only 1.4%16.

Recent studies have shown the potential increased diagnostic value of three-dimensional ultrasound in the prenatal evaluation of facial clefts, particularly of the palate, compared with two-dimensional ultrasound, as evaluation of the secondary palate is possible. Three-dimensional ultrasound provides a clearer and more precise visualization of the fetal primary and secondary palate, allowing differentiation of the position and extent of the cleft, especially in cases in which 2D ultrasound is limited by acoustic shadowing6, 7. A new approach incorporating the evaluation of the retronasal triangle into first-trimester screening is of potential value in the early detection of cleft lip and palate8.

MRI is a rapid and powerful tool for the identification of fetal malformations and can be used as a secondary imaging technique when ultrasound is inconclusive17. MRI is less affected by maternal body habitus, severe oligohydramnios and unfavorable fetal position than is ultrasound examination, and the soft-tissue resolution is high18. Fetal MRI provides a detailed evaluation of the fetal facial anatomy, as well as the presence of associated anomalies and can be used as an imaging adjuvant in doubtful cases to confirm the position and extension of clefts in patients at risk.

There are significant clinical differences in prognosis depending on the degree of extension of the cleft. If the cleft involves the palate there are severe implications for feeding, hearing and maxillary development19. Precise classification of the fetal malformation is important as it offers the parents the possibility of antenatal consultation with members of the craniomaxillofacial treatment team, focusing on possible developmental problems and the extent of surgery required after birth20.

This study is the first to include a large series of cases of cleft lip and palate detected by MRI. Only one prospective study investigated and showed the usefulness of fetal MRI in the assessment of facial clefts; Ghi et al.5 confirmed the presence of a facial cleft in only six cases at a mean gestational age of 30 weeks. Our study shows that the same information can be gathered in even younger fetuses. Twenty (58.8%) cases in our study were imaged before 26 weeks of gestation, and two were imaged as early as 19 weeks of gestation. The location of the cleft, its extension into the secondary palate and the degree of the extension were correctly identified in all 30 fetuses with complete follow-up.

Of note, in our study population, sonography was performed as a screening modality and MRI as a secondary imaging technique in patients at risk. This must be considered when MRI results are compared to ultrasound diagnosis. Nevertheless, our findings suggest that MRI could be useful in identifying and locating facial clefts when ultrasound is limited, and in increasing confidence in the ultrasound diagnosis.

The additional knowledge provided by MRI might improve prenatal parental counseling and postnatal therapeutic planning by the cleft lip and palate team. It offers precise visualization of the fetal palate and allows evaluation of the laterality of the cleft and the extension of the malformation into the secondary palate. When in-utero endoscopic techniques21 become a reality in prenatal surgery, repair of facial clefts should prove to be the optimal therapy as fetal facial wounds may heal in utero without scarring22, 23, making the precise prenatal definition of the type and extension of a facial cleft even more important in the near future.

Ancillary