Accuracy of Prenatal Ultrasound Scans for Screening Cleft Lip and Palate

A Systematic Review

Abstract

Children born with clefts encounter various postnatal issues which have a negative impact and long-term psychological effects on both the parents and themselves. This systematic review aims to find the accuracy of two-dimensional (2D)/3D scans for screening cleft lip and palate which would help the parents to be mentally and psychologically prepared to plan for future surgeries. To identify relevant literature, an electronic search was performed using PubMed, Trip database, Cochrane, and Google Scholar database. The search yielded 927 articles, of which 10 fulfilled the eligibility criteria and were included in this systematic review. These studies showed that the combination of 2D and 3D ultrasound scans has increased the specificity and sensitivity in detecting orofacial clefts, as they aid in improved visualization of the fetal face. However, in future, the visualization of the palate can be improved by inventing other modified views as cleft palate diagnosis is more prone to error.

INTRODUCTION

Cleft lip and cleft palate are the most common craniofacial birth defect affecting approximately 1–1.5/1000 live births worldwide and between 27,000 and 33,000 clefts per year in India.[¹] The cleft children encounter various postnatal disputes that negatively impact their quality of life.[²]

Hence, detection of these clefts using antenatal two-dimensional (2D)/3D ultrasound scans prepares the parents psychologically to plan for postnatal surgery.[³⁴] Since there is limited information on the sensitivity and specificity of these ultrasound scans, this systematic review aims to identify the accuracy of prenatal ultrasonographic detection of cleft lip and palate.

MATERIALS AND METHODS

To obtain, evaluate, and summarize all relevant findings relating to the accuracy of prenatal ultrasound scans for screening cleft lip and palate, a systematic review was conducted as per the PRISMA guidelines. The International Prospective Register of Systematic Reviews was used to register the protocol for this systematic review (ID: 307957).

Eligibility criteria

The PECO analysis of the articles searched was shown below:

• Population: Pregnant women who have undergone prenatal screening
• Exposure: Prenatal 2D/3D ultrasound scans
• Comparison: Not applicable
• Outcome: Detection rate of fetal cleft lip/palate.

Inclusion and exclusion criteria

Inclusion criteria

1. Cross-sectional, retrospective, and prospective cohort study designs
2. Studies done in the period from January 2012 to January 2022
3. Studies in which children were screened for cleft lip only (cleft lip without cleft palate), cleft lip with or without cleft palate, or cleft palate only (cleft palate without cleft lip) and children having syndromic cleft lip and palate were included
4. Studies that were written in the English language were only included.

Exclusion criteria

1. Studies that assessed screening other than ultrasound like computed tomography and magnetic resonance imaging were excluded
2. Qualitative studies, reviews, expert opinions, systematic reviews, meta-analyses, and case studies/series
3. Publications with no abstract, conference papers, editorials, and those which were widely out of the scope of the study were excluded.

Search strategy

A broad search of the literature was performed in PubMed, Trip database, Cochrane, and Google Scholar database was performed in the specified time period. The search strategy included the combination of the following terms: “cleft lip,” “cleft palate,” and “ultrasound.” The MeSH terms used in PubMed were “cleft lip,” “cleft palate,” and “ultrasound.” The bibliographies of the selected articles were also searched to ensure that no relevant articles were missed. The relevant studies were sorted out on the basis of their title and abstract. Finally, those studies during which the abstract fulfilled all the inclusion criteria were selected for full-text reading. Further, the full text of all the articles whose abstracts met the inclusion criteria was read. The search strategy for various databases is shown in Table 1. Figure 1 shows the PRISMA flowchart.

Figure 1: Preferred Reporting Items for Systematic Reviews and Meta-Analyses flowchart

Table 1: Search strategy for systematic review

Data extraction

Data extraction from the ten included articles was done using a data extraction form. It included the first author's name, year of publication of the article, study design, gestational age of screening in weeks, 2D/3D ultrasound scans, features, outcome, and inferences. Table 2 presents the information extracted from the studies included during this systematic review.

Table 2: Study characteristics

Quality assessment of the included studies

The final analysis included ten cohort studies. The tool used for quality assessment is by Newcastle[FIGURE DASH]Ottawa Quality Assessment Form for cohort studies.[⁵] The methodological quality of the selected articles which was assessed using the Newcastle[FIGURE DASH]Ottawa Form for cohort studies is summarized in Table 3. For cohort studies, the quality score was based on the following categories: selection, comparability, and outcome. A study is often awarded a maximum of one star for each item in the selection and outcome categories. A maximum of two stars are often given for comparability. Stars have been converted into scores for our convenience. The higher the score, the better the quality of the study. The studies with scores 9–7 are good studies, 5–6 scores are satisfactory studies, and 0–4 scores are unsatisfactory studies. Further, these scores have been categorized into high, unclear, and low risk of bias using Review Manager Software version 5.3. Bias risk assessment results are shown in Figure 2, and the risk of bias summary is shown in Figure 3.

Figure 2: Bias chart

Figure 3: Risk of bias summary

Table 3: Quality assessment of the included studies using the Newcastle?Ottawa Scale for cohort studies

RESULTS

Search results

The search generated a total of 943 articles based on the title from four different electronic databases: PubMed, Trip database, Cochrane, and Google Scholar. PubMed produced 627 articles, Trip database produced 268 articles, Google Scholar produced 36 articles, and 12 articles were found in Cochrane. Among the obtained articles, 16 articles were eliminated due to duplication, 915 articles eliminated after abstract reading, and 2 articles eliminated after full-text reading. Finally, ten articles were selected for the review using a PRISMA flowchart based on inclusion and exclusion criteria. The inter-examiner bias was eliminated. The first two authors (DK and IS) analyzed the studies. The discrepancies between the first two authors were resolved by the consensus from the third author (MK). Out of these, seven articles were rated as good quality, three with satisfactory quality, and none of the articles were rated as poor quality.

Outcome

A study done by Berggren et al. on 2D or 3D scans found that the sensitivity of clefts was 43% and specificity of 100% excluding isolated cleft palate. However, additional scanning views of the face need to be followed in those cases in which the fetal face seemed to be impossible to visualize.[⁶] Loozen et al. studied clefts using both 2D and 3D scans. His findings reveal that out of 76.9% of cases, there was an underestimation of clefts in 19.4% and overestimation in 3.7% of cleft cases. 3D reverse face view allows a relatively straightforward evaluation of the fetal palate with a high degree of accuracy.[⁷] Gindes et al.'s study on 3D ultrasound scans on cleft palate found the sensitivity, specificity, positive predicted value, and negative predicted value to be 71.4%, 91.9%, 62.5%, and 94.4%, respectively. Acoustic shadows must be avoided, and at least two different angles of 3D imaging are required for a better diagnostic accuracy of cleft palate.[⁸] Studies conducted by Martinez-Ten et al. using 3D ultrasound scans found that all cases had clefting of the primary palate and 86% of clefts in the secondary palate with a satisfactory false-positive rate. Identification of retronasal triangle in the coronal plane recognizes the primary and secondary palate of a cleft fetus, as it has a resonable degree of certainity.[⁹] A study done by Lam et al. found that the overall accuracy of clefts visualized on 2D and 3D scans was 95%. Flipped face and oblique face views allow a higher chance of satisfactory visualization of the secondary palate.[¹⁰] The study conducted by Venezia et al. on 2D and 3D scans found the sensitivity of 2D and 3D to be 50% and 100%, respectively.[¹¹] Lakshmy et al.'s study on 2D and 3D scans found no false-positive results with one missed bifid uvula case. Although 2D axial, sagittal, and coronal views are complementary to one another in detecting cleft palate, a further 3D evaluation would be necessary.[¹²] Faure et al. conducted a study on 2D and 3D ultrasound scans on cleft palate, which found the adjusted kappa coefficient between prenatal and postnatal findings to be 0.88 which corresponds to an excellent agreement. 3D imaging using modes such as tomographic ultrasound imaging or volume contrast imaging offers a realistic visualization of the cleft palate.[¹³] The sensitivity of 2D and 3D scans with reformatting technique by Ji et al. was found to be 80% and 92%. 3D ultrasound reformatting technique using Omniview displays the full extent and severity of the secondary palate, and it could also overcome the shortcomings of other 3D ultrasound technologies.[¹⁴] Deng et al. found the sensitivity of clefts on 2D and 3D scans to be 36.8% and 89.5%, respectively. 2D ultrasound is mainly affected by the fetal position along with the scanning angle. Hence 3D imaging analyses cleft lip and palate with good reproducibility even at different angles and directions of cleft sites.[¹⁵]

DISCUSSION

Cleft lip/palate is due to the disturbance in the growth and fusion of five facial prominences between the 4^th and 10^th weeks of human intrauterine period of development. In the 5^th week, two bilateral nasal pits divide the frontonasal prominence into lateral and medial nasal processes. By the end of the 7^th week, the intermaxillary segment is formed which fuses along the midline together with the maxillary prominences from posterior to anterior. Failure of this development results in the cleft of the lip, alveolar process, and hard palate anterior to the incisive foramen (primary palate). The secondary palate which is located posterior to the incisive foramen is fused by the lateral palatal shelves of the maxilla prominences. Failure of this fusion results in the cleft of the secondary palate.[¹⁶]

Children born with cleft lip and cleft palate are at a greater risk of feeding difficulties, language disorders, ear infections, anatomical anomalies, and other postnatal disputes. Families of these cleft children also experience long-term psychological effects.[¹⁶] Hence, prenatal ultrasound has been utilized to diagnose these clefts using 2D and 3D methods at an early stage so that the parents would be mentally prepared and plan for postnatal surgeries.[¹⁷] The accuracy of ultrasound diagnosis of clefts varies widely and has increased over the years with the improvement of ultrasound equipment. With standard sonography, these facial clefts can be classified by a combination of coronal, sagittal, oblique, and axial scansions.[¹¹] Hence, this systematic review aims to find the accuracy of those prenatal ultrasound scans in detecting cleft lip and palate children.

The diagnosis of cleft lip and palate in the first trimester has been scarcely reported in the literature. Martinez-Ten et al. in 2012 conducted a study using 3D ultrasound at the time of 11 to 13 weeks of gestation. The occurence of some false-positive cases in the study illustrated that early prenatal diagnosis of cleft lip and palate has certain limitations. However, a dedicated second-trimester scan for evaluation of the lip and palate should be mandatory in those cases.[⁹]

In 2012, Berggren et al. have concluded that the rate of identifying the facial clefts can be improved by focusing on the standardizing screening plans and rescanning if incomplete facial view is obtained.[⁶]

The accuracy of ultrasound in determining the type of cleft is not sufficient in much research, and in order to overcome this, Loozen et al. in 2015 had conducted a study using both 2D and 3D scans and found that no errors were made in distinguishing unilateral from bilateral clefts. This study is a reliable method to find the type of cleft though cleft palate is easily missed and underestimation of the extent of the cleft is common.[⁷]

3D ultrasound has an advantage over the 2D ultrasound in the spatial orientation and the third planar image and improves the interpretation of 2D images of clefts. Gindes et al. in 2013 had conducted a study on visualizing cleft palate in 2D and 3D scans at 12-40 weeks of gestation. He had found that the misinterpretations were at the second and third trimesters when acoustic shadows are more dominant. The cases during which the palate could not be demonstrated were within the third trimester. In most high-risk cases, detection of the fetal cleft palate during pregnancy is possible with an accuracy of about 90%.[⁸]

The overall detection rate of facial clefts by midtrimester ultrasound was comparable to that reported in the literature. Lam et al. in 2015 evaluated the accuracy of 2D and 3D scans before 24 weeks of gestation. He had found that the most common inconsistency was in the overdiagnosis or underdiagnosis of alveolar clefts, whereas there have been no errors concerning the side of the cleft. The limitations of ultrasound predictions should be explained to parents at the time of counseling.[¹⁰]

Systematic screening 2D ultrasound requires at least two scans. The median sagittal scanning allows visualization of the fetal profile and the philtrum protruding. Fore coronal scan of the nose and mouth allows the display of the upper lip and the possible alignment abnormalities of the alveolar processes. Venezia et al. in 2016 demonstrated the reliability of the screening of the midtrimester in the diagnosis of facial clefts with 2D and 3D ultrasound concluding the sensitivity of 2D ultrasound to be 50%, while the 3D ultrasound was found to be 100%. Combining ultrasound 2D + 3D is the one that guarantees the best results of sensitivity and specificity.[¹¹]

Screening for palatine clefts starts with a good 2D assessment of bony landmarks of the palate. The sagittal, axial, and coronal views are complementary to every other in suspecting cleft palate (CP), which might need further 3D evaluation. Lakshmy et al.'s study in 2017 demonstrated that the evaluation of the palate can be done at the 11–14-week scan based on 2D markers and can also be assessed with 3D sonography. The inclusion of 2D markers in all three planes increases the sensitivity for the detection of palatine clefts.[¹²]

Faure et al. in 2020 had conducted a study by visualizing the fetal secondary palate in 3D plane, starting with 2D axial transverse ultrasound view. He had found the adjusted kappa coefficient between prenatal and postnatal evaluation to be 0.88 (95% confidence interval: 0.79-0.97) corresponding to an excellent agreement. Using a strictly axial transverse ultrasound view, visualization of the secondary fetal palate enables to diagnose cleft palate without cleft lip. This method offers a prenatal anatomic classification of cleft palate with a high level of concordance to postnatal findings.[¹³]

Reformatting technique such as Omniview can enhance the detection of isolated cleft palate which is difficult to visualize in traditional ultrasound. Any line, curve, or polyline can be drawn with this technology along the structure of the palate to display the entire palate in a single image. Ji et al. had conducted a study with both 2D and 3D with reformatting technique and found that 3D with reformatting technique has high diagnostic accuracy for cleft lip and alveolus. However, 3D ultrasound has a much higher diagnostic accuracy for cleft lip and palate.[¹⁴]

Currently, 2D ultrasound has been able to successfully diagnose cleft lip deformity, and 3D ultrasound as a screening test has proven to be a feasible screening tool. Deng et al. had concluded that there was no significant (P > 0.05) difference in the 2D and 3D ultrasound detection rate of the pure cleft lip. 3D ultrasound can show the structure of the palate that is difficult to display with conventional 2D ultrasound, which significantly improves the detection rate of prenatal fetal harelip and palate, especially in the diagnosis of cleft palate, which has greater advantages than 2D ultrasound examination.[¹⁵]

In this review, only those articles published in the English language were included. Additional records identified through gray literature were not included. All studies comprised in this review were cohort studies, and further, more longitudinal studies are required to have concrete evidence on the accuracy of 2D and 3D scans in detecting children born with cleft lip and palate.

The success rate of ultrasound scans depends mainly on the fetal position and on the experience of the sonographers. In future, the limitation in the visualization of the palate can be improved by inventing other modified views as cleft palate diagnosis is more prone to error.

CONCLUSION

2D and 3D ultrasound scans have the same accuracy for cleft lip. However, if a cleft lip is suspected, 3D ultrasound should be used for secondary evaluation after the 2D ultrasound. 3D imaging is simple to implement and it improves the prenatal detection rate of cleft lip and palate, especially those involving the secondary palate. This 3D imaging from different directions and angles of cleft lip and palate was analyzed with good reproducibility. In conclusion, the combination of 2D and 3D ultrasound scans is the best approach to increase both the specificity and sensitivity as it aids in improved visualization of fetal face.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.