Prenatal ultrasound examination of the secondary palate

Two papers published in this issue of the Journal make significant contributions to the prenatal diagnosis of clefting of the secondary palate. The secondary palate consists of a hard palate, which runs posterior and horizontal to the incisive foramen, and a soft palate or velum, which curves downwards and backwards from the posterior aspect of the hard palate and ends in the uvula. In the fetus, the relative length of the hard to the soft palate is 2 : 1; the soft palate is fleshy and of a similar thickness to that of the bony hard palate (Figure 1). Clefting of the secondary palate is always midline and results from failure of the palatine processes to elevate and grow. It is usually associated with clefts of the lip and alveolus, but isolated clefts of the secondary palate account for 25–80%, depending on the geographical location, of all orofacial clefting. Clefting of the hard palate always involves the velum, but it is possible for the soft palate to be cleft in association with an intact hard palate. Lesions involving the hard palate can require complex surgery and are associated with difficulties of speech, hearing and feeding, and (after surgery) mid-face protrusion. Isolated clefts of the soft palate are easier to treat surgically, but are often missed at the neonatal examination. Anatomical section of a 24-week fetus showing a sagittal view of the nasal cavity (septum intact), lips, alveolus, secondary palate, tongue and oral cavity. The soft palate is between the arrows. The vomer bone is outlined. Courtesy of Dr Gonzalo Moscoso with thanks. Before the advent of three-dimensional (3D) ultrasound, the rate of successful diagnosis of lesions of the secondary palate was extremely low. Clefts of the lip and primary palate could be diagnosed by the two-dimensional (2D) oblique face view1, 2, but severe shadowing by the maxillary bone made visualization of the secondary palate virtually impossible. Nevertheless, in an interesting paper, Sherer et al.3 claimed that visualization of the secondary palate in the axial plane by 2D was not difficult, and produced the only existing nomograms of the length, breadth and area of the hard palate. This group did not, however, present any cases of defects of the secondary palate diagnosed by this method. Initial studies with 3D ultrasound by both multiplanar reconstruction and surface rendering were found to be helpful in the visualization of clefts of the lip and primary palate by several groups4-8, although in none of these studies was there a systematic attempt to diagnose clefts of the secondary palate. Rotten and Levaillant9, in a retrospective series of 96 cases of orofacial clefting, attempted to categorize the defects with respect to whether they involved the lips, alveolar ridge or secondary palate. They observed that the alveolus and maxilla were best seen in an axial plane and the secondary palate in a coronal plane. Although there was 87% concordance between the sonographic report and postnatal evaluation, five secondary palatal defects were missed and there were four false-positive diagnoses of defects in the secondary palate in fetuses with clefts of the lips and alveolus. The basic problem was acoustic shadowing of the palate by the maxilla when front-to-back scrolling was carried out in the coronal plane. Campbell et al.10 overcame this problem by rotating the face through 180° and scrolling from back to front. This technique, described as the ‘reverse-face view’, eliminated shadowing by the maxilla and good coronal views were obtained of the normal palate and clefts of the secondary hard palate. Shadowing could occur if there was a limb in front of the face, but the authors regarded this as a temporary problem. Cord and placenta did not obscure the view of the palate. Degrees of obliquity of more than a three-quarter profile were unsuitable for obtaining a good reverse-face view. In their series of fetuses with orofacial clefting, all seven cases of clefting of the hard palate, ranging from 23–34 weeks' gestation, were diagnosed correctly. The remaining fetus had an intact hard palate, which was categorized correctly, but a cleft in the soft palate was missed. The authors felt that the method was rapid and reliable but that there may be difficulties in visualizing the soft palate, as the velum hangs down almost vertically and would not be at right angles to the beam in a coronal section. Benacerraf et al.11 described the diagnosis of an isolated cleft of the soft palate from an axial 3D view of the palatal area. However, the lesion was shown clearly on coronal magnetic resonance imaging, meaning either that coronal sections may be satisfactory to make this diagnosis or that the lesion extended into the hard palate. The defect did not have sharp edges on ultrasound examination. Platt et al.12 described the ‘flipped-face view’ to obtain 3D rendered axial sections of the hard and soft palates. A very detailed description was given on how to obtain views of the inferior surface of the palate, scrolling from chin to nose. Using this technique, the mandible, tongue, maxilla and palate were examined systematically and images of normal palatal anatomy and clefts of the lips, primary and secondary palate were presented. The authors found this technique to be easy to perform but pointed out the dangers of a false-positive diagnosis of clefting due to shadowing of the palate by the maxilla. Care should be taken to make minor rotational adjustments to avoid this problem. The authors believed that the technique provided views of the soft palate, but in the image presented there was no clear evidence that the structure identified as the velum was not in fact the posterior part of the hard palate, especially as there appeared to be an interpalatal suture, which is typical of the hard palate. A case of orofacial clefting examined by this method demonstrated a cleft in the palate but once again the lesion did not have sharp edges. The authors examined 50 cases of facial clefting by this method and postnatal examination showed 100% concordance with prenatal findings. The gestational ages at which the examinations were made was not given. The paper by Faure et al.13 in this issue further examines the role of the axial plane in visualizing the secondary palate on 3D multiplanar imaging. The authors were able to obtain satisfactory views of the palate in all 100 low-risk cases, all of which had normal anatomy. Examinations were performed at 17, 22, 27 and 32 weeks' gestation and images were compared with pathological fetal specimens at the same gestational age. The images of the palate, especially their Figures 2a (22 weeks) and 3b (27 weeks) were strikingly good and anatomical features such as the alveolar ridge with tooth buds, the palatine process, the interpalatal suture, the horizontal plate of the palatine bone, the transverse palatine suture, and the posterior nasal spine and pterygoid processes could be identified clearly. The interpalatal suture became less identifiable with increasing gestational age, but the alveolar ridge and horizontal plate were always identified. Surprisingly, the authors do not report difficulties with shadowing of the palate by the maxilla. However, what is evident from the remarkable detail provided of the posterior edge of the hard palate is that the soft palate was not visualized. Multiplanar and surface-rendered images of a 16-week fetus showing normal nose and lips with axial views of the superior and inferior aspects of a normal hard palate. The paper by Pilu and Segata14, also in this issue, describes a new multiplanar approach to the diagnosis of clefts of the secondary palate. Recognizing that acoustic shadowing by the alveolar ridge was a potential problem when it comes to identifying the hard palate, the authors insonated the secondary palate at an angle of 45° and obtained satisfactory views of the secondary palate in 10 of 15 cases between 19 and 28 weeks. Failure was always the result of an unfavorable position of the fetus. The authors felt that surface rendering was not required for adequate visualization. Tomographic images of the palate in the coronal plane were also obtained using the TUI modality. Although part of the soft palate could be visualized in the sagittal plane, it was not demonstrated in the axial and coronal planes. In a single case of orofacial clefting the axial plane was not fully informative due to the tongue masking the edges of the lesion. The coronal planes displayed the defect clearly. Studies of the secondary palate can be made in the second and third trimesters by means of 3D multiplanar or TUI scanning. The use of surface rendering modalities appears to improve visualization. The palate can be seen in the coronal plane by the ‘reverse-face view’ and the axial plane by the ‘flipped-face view’. Both views can be affected by acoustic shadowing, but it is more problematic in the axial plane; this technique is therefore probably better suited to second-trimester diagnosis. Axial views provide detailed anatomical information of the inferior surface and the posterior border of the hard palate. Coronal views give the relationship of the palate to the mouth and tongue inferiorly and the nose and orbits superiorly. The precise location of the section of the palate on a coronal view can only be made by reference to the sagittal section using the multiplanar or TUI modality. There is no convincing evidence to date that the soft palate can be visualized on either coronal or axial sections of the palate. Although this is disappointing, isolated soft palate lesions are easier to correct surgically. Clefts of the hard palate are probably better seen on coronal sections where the relationships of the edges of the defect to the tongue are usually well-demonstrated. To date, few examples of clefting on axial sections have been published, but the edges of the lesion seem less well-defined. In the axial plane care must be taken to avoid false-positive diagnoses of clefting due to acoustic shadowing by the maxilla and false-negative diagnoses by mistaking the tongue for an intact palate. The anatomical landmarks described in recent papers will help to avoid this. Fetus at 16 weeks; after examination of the face, a surface-rendered view of the secondary hard palate is obtained in the axial plane by the ‘tilt and scroll’ method. Fetus at 13 weeks; after examination of the face, the hard palate is examined in the axial plane by the ‘tilt and scroll’ method.