Facial profile markers during first trimester is composed of facial bones. The development of facial bones is related to the migration and differentiation of neural crest cells from ectoderm, and presents a certain sequence[12]. At the 4th week of embryonic development, there are five protuberances around the primitive oral cavity: the frontonasal protuberance, the paired mandibular protuberances and the maxillary protuberances[13]. CLP is due to the growth and fusion disorders of these five protuberances[14]. The ossification of maxilla and mandible begins from 8 weeks onward[15]. During the first trimester, the mandible is indirectly connected to the skull through the temporomandibular joint and muscles, and the forward growth rate is not as fast as that of the maxilla, which is directly connected to the skull[16, 17]. From 20 weeks of gestation, the development of fetal swallowing function accelerates the growth of the mandible, while the maxilla ossification has initially completed. Subsequently, the position of the facial bones is relatively constant, and the fetal facial profile is basically formed[17]. In this study, multiple facial markers reflecting the relative position of forehead, maxilla, nasion and mandible were selected to analyze their early diagnostic value for trisomy 21, trisomy 18 and CLP fetuses during the first trimester.
In 2002, Rotten et al. [4] first introduced a quantitative facial angle called IFA to describe the relative position between mandible and frontal bone. They found that IFA was below the 5th percentile in 25% of trisomy 21 fetuses during the second and third trimester. When it was below 49.2° (average-2SD), micrognathia could be suspected, with a sensitivity of 100%, a specificity of 98.9%, a positive predictive value of 75% and a negative predictive value of 100%. IFA was mainly used to objectively evaluate the micrognathia, so as to indicate Pierre-Robin syndrome, Stickler syndrome, trisomy 18 and trisomy 13[5, 18, 19]. In our study, the IFA values of trisomy 21 and trisomy 18 fetuses were significantly smaller than the normal group. Moreover, the diagnostic accuracy of trisomy 18 was much higher than trisomy 21 (83.1%>69.4%). IFA below 75.37° during the first trimester was significant for suggesting trisomy 18, with a sensitivity of 70.4% and specificity of 85.7%. This cut-off was much larger than Rotten et al. [4] (49.2°), which might be due to the slower forward growth rate of mandible than maxilla during the first trimester. However, the clinical significance of the cut-off need to be confirmed by a large sample study.
The MNM angle could reflect the relative position of maxilla, nasion and mandible. Bakker et al. [10] reported the MNM angle abnormally increased (above the 97.5th percentile) in bilateral CLP and micrognathia during the first trimester, confirming the predictive value in abnormal facial profile and special facial malformations. According to Lu et al. [7], in 4 cases of trisomy 18 with micrognathia, the MNM angle values were above the upper limit of normal values, which were similar to the results of de Jong-pleij et al. [6]. Similarly, in our study, the MNM angle was above the 95th percentile of normal values in 42.9% of trisomy 18 and 80% of CLP fetuses. However, it had an accuracy of 96.4%, a sensitivity of 91.2% and a specificity of 90% to detect CLP, which was superior to detecting trisomy 18. Therefore, we boldly speculated that the MNM angle was a sensitive indicator for judging facial profile during the first trimester. When it abnormally increased and was above 6.83°, the possibility of CLP should be considered. Meanwhile, we found that there was no statistically significant difference in MNM angle between trisomy 21 and the normal group. It might be related to the fact that most trisomy 21 fetuses were accompanied by midfacial hypoplasia, low tongue tension and outside oral cavity [20]. Another possible explanation was that trisomy 21 had hypoplasia of the maxilla as well as hypoplasia of the mandible to a less degree [21, 22]. In that case, the position of the maxilla and mandible was changed, but the angle between them (MNM angle) might remain normal.
FMA could directly reflect the relative position of maxilla and mandible.
Lu et al. [7] observed that the variation of the value of FMA changed very little after 16 weeks’ gestation (the minimum was 64.6° at 16 weeks, and the maximum was 67.1° at 28 weeks). Using 66° as a cut-off, FMA had a detection rate of 100% for micrognathia (the false positive rate as low as 2.5%). In our study, the FMA values of trisomy 18 were significantly smaller than the normal group. The cut-off of FMA in diagnosing trisomy 18 was 67.96°, with a sensitivity of 85% and a specificity of 85.7%. The value was similar to the results of Lu et al [7]. A large prospective cohort was needed to confirm the accuracy of FMA in the diagnosis of trisomy 18 during the first trimester.
The FS distance could be used to evaluate the relative position of maxilla, mandible and frontal bone on the basis of MML. Yazdi et al. [8] showed that when FS distance was added into the first-trimester combined screening for aneuploidy, the false positive rate could be reduced from 5–3% while the detection rate of aneuploidy (90%) remained unchanged. In 2016, Hoopmann et al. [9] pointed out abnormal FS distance could detect some tiny maxillary protrusion, which had a definite implication for fetal CLP during the first trimester. Similarly, we found that the FS distance of CLP was significantly smaller than the normal group. Setting the cut-off to 0.35mm, the FS distance had an accuracy of 89%, a sensitivity of 86.8% and a specificity of 80% to diagnose CLP. Once abnormal FS distance was found, it was necessary to pay attention to whether the fetus had facial malformations (such as CLP) and chromosomal abnormalities.
The PL distance was affected by the position of mandible, nasion and frontal bone. Jong-pleij et al. [23] pointed out that the PL distance above 4mm was an objective indicator to predict frontal bossing during the second and third trimester. Nevertheless, in our study we concluded that the PL distance of trisomy 21, trisomy 18 and CLP fetuses had no significant difference from the normal group. Therefore, PL distance was not the best ultrasound markers for trisomy 21, trisomy 18 and CLP during first trimester. This is consistent with Bakker et al. [10]. However, we need a large number of prospective studies to verify it.
The main limitation of our study is the number of abnormal cases. It’s a bit less. A large multi centered prospective cohort study is required. Secondly, 6 cases of trisomy 21 and 1 case of trisomy 18 fetuses had absent or shortened nasal bone, and the other 14 cases of trisomy fetuses showed no obvious facial malformations. Thirdly, chromosomal analysis was not performed in 10 cases of CLP fetuses due to personal reasons, which limited the analysis of results to a extent. Additionally, 2D ultrasound images were used in this study. Some scholars considered that three-dimensional (3D) ultrasound could obtain a real standard mid-sagittal section through multi-plane mode [24]. However, numerous researches show that 2D ultrasound is the basis of 3D ultrasound reconstruction imaging and 10% of 3D reconstructed images could not be used for NT evaluation [25]. As a consequence, the measurement of facial markers by 2D ultrasound was easier for clinical application. Lastly, it was worth mentioning that the measurement of these facial markers was time-consuming and complicated. In the future, we will integrate artificial intelligence (AI) with these markers to promote the establishment of intelligent medical mode and increase the efficiency.