Systematic Ultrasound Evaluation of Olfactory Sulci in Fetuses with Congenital Heart Defects: A Clue for CHARGE Syndrome Diagnosis

The prevalence of congenital heart defects (CHDs) in liveborn neonates is 1%, and they are associated with high morbidity and mortality rates [1, 2]. Although most CHDs have a multifactorial etiology, several well-known genetic syndromes are linked to CHDs and should be considered during prenatal evaluation of fetal cardiac anomalies [1‒3]. Prenatal exome sequencing (ES) currently provides a diagnosis in approximately 30% of structurally abnormal fetuses with nondiagnostic chromosomal microarray (CMA) or karyotype [4]. However, it is not yet considered a standard part of prenatal evaluation for CHDs. To improve the diagnostic rates of genetic syndromes associated with CHDs, efforts are needed to achieve adequate phenotyping of the prenatal imaging studies [2, 5]. This is particularly important in monogenic diseases such as CHARGE syndrome [6] (OMIM #214800), which has a significant impact on postnatal morbidity and is typically not detectable by standard CMA test [7]. CHARGE syndrome is a rare genetic syndrome primarily caused by a heterozygous mutation in the CHD7 gene on chromosome 8q12, with an estimated incidence of 1 in 10,000 births [8]. The term “CHARGE” is an acronym that describes a constellation of clinical features, including Coloboma, Heart defects, Atresia choanae, Retardation (of growth and/or development), Genitourinary malformation, and Ear abnormalities [6]. Since some of the major postnatal diagnostic criteria of CHARGE syndrome [9, 10] are rarely seen prenatally, the major diagnostic criteria of CHARGE syndrome during fetal life differ from the strict postnatal criteria and include arhinencephaly [11]. Arhinencephaly refers to the agenesis of olfactory bulbs and tracts and is characterized by congenital anosmia, present in almost all cases of CHARGE syndrome [12]. Arhinencephaly is also related to the abnormal development of olfactory sulci (OS) [11, 13, 14] which are orbitofrontal structures separating two horizontal cortical gyri: the rectus gyrus medially and the orbital gyrus laterally [14]. OS can be evaluated by fetal ultrasound [13] and has been identified as a key feature of CHARGE syndrome [11, 12].Given that central nervous system (CNS) imaging is now mandatory for fetuses with CHD [15, 16], we hypothesize that assessment of the OS should be incorporated into routine neurosonography (NSG) for this population. The objective of this study was to evaluate, for the first time, the systematic assessment of the OS in fetuses with major CHD through NSG and to investigate its potential contribution to the diagnosis of CHARGE syndrome in this population.

This prospective study included all fetuses with a prenatal diagnosis of major CHD who were followed up in a tertiary center (BCNatal – Hospital Clínic and Hospital Sant Joan de Déu) from January 2017 to December 2021. All included cases underwent detailed fetal echocardiography performed by a fetal medicine specialist with expertise in fetal cardiology to confirm the diagnosis. NSG was then performed between 30 and 40 weeks of gestation. Cases in which NSG was not performed due to pregnancy termination, premature labor, or intrauterine fetal demise were excluded from the final analysis. Additionally, cases with incomplete imaging (specifically, those lacking trans-coronal plane images) and those in which CHD was not confirmed postnatally were also excluded from the analysis. Gestational age calculation in all pregnancies was based on the crown-rump length at first-trimester ultrasound [17]. After a case was classified as having abnormal OS, a T2-weighted magnetic resonance imaging (MRI) scan, using single-shot fast spin echo, was performed and interpreted by a radiology specialist at our diagnostic imaging department. MRI was performed to confirm the diagnosis of the OS anomaly and to further explore additional structural anomalies associated with CHARGE syndrome that are difficult to assess via ultrasound, such as choanal atresia, colobomas, and semicircular canal anomalies. It should be noted that the radiologist was not blinded to the previous NSG results. In addition, a trio-based clinical exome sequencing (CES) test was performed after confirming normal CMA results during genetic counseling. The diagnosis of CHARGE syndrome was confirmed through genetic testing. The study was approved by the Institutional Ethics Committee (HCB/2015/0365), and all patients provided written informed consent.

Advanced NSG always included a three-plane evaluation using the transabdominal approach and, when possible, the transvaginal approach. In all cases, NSG was performed by a fetal medicine specialist expert in fetal NSG (E.E., M.P.C., M.I., N.M., and E.M.) using the GE Voluson S8 or GE Voluson E10 expert ultrasound system (GE Healthcare, Zipf, Austria). OS were identified in the trans-frontal coronal plane, following the guidelines of the International Society of Ultrasound in Obstetrics and Gynecology [15, 16]. Table 1 presents the three key planes of NSG along with the corresponding anatomical landmarks required for each plane. The image selected for OS assessment had to show a clear visualization of the landmarks described by Acanfora et al. [13]: the interhemispheric fissure, frontal lobe periventricular echogenicity, and the sphenoidal bones with the orbital ridge (Fig. 1). OS appearance was classified as: fully developed when deep hyperechogenic lines were seen, underdeveloped when there were only smooth, shallow depressions, and absent when the cortical surface was smooth without any observed depressions [13]. All images were stored for later offline evaluation by another independent and blinded examiner. A complete fetal anatomy revaluation was done when abnormal OS was found.

A total of 211 fetuses with CHD were evaluated in the fetal cardiology unit during the study period. Among them, 147 fetuses with major CHD with both normal and abnormal chromosomal microarray test results (8 cases) were included in the study. Cases where NSG was not performed or was incomplete and cases without stored images were excluded. CHD was confirmed postnatally in all cases. The flowchart of eligible cases is depicted in Figure 2. The mean gestational age at the time of NSG was 34 ± 2 weeks (mean ± SD), and evaluation of OS existence and the developmental stage was successful in 143 cases (97% of cases). The mean maternal age was 33 ± 6 years (mean ± SD), with 48 (33%) nulliparous pregnancies. Fetal gender distribution was 85 (58%) females and 62 (42%) males. The distribution of the different types of CHDs in the study population is presented in Table 2. OS appeared symmetric in all cases, and no classification discrepancies were found between the different examiners. At the time of the first NSG, abnormal OS was found in four fetuses (2.7%): three with no visible OS and one with hypoplastic OS (Fig. 3). A detailed description of all 4 cases is presented in Table 3. In cases 1 and 4, the pregnancies were terminated at the parents request after genetic confirmation of CHARGE syndrome. Three fetuses were small for gestational age (SGA) at the end of the pregnancy. In case 3, preterm delivery occurred at 35+3 weeks due to preterm contractions, and OS were eventually visible as significantly underdeveloped only on the day of delivery. Wide perimembranous ventricular septal defect was the CHD in all the cases with abnormal OS. Prenatal MRI confirmed OS and olfactory bulbs abnormality in all 4 cases evaluated by NSG (Fig. 4). Normal cochleovestibular structures were not seen in any of the cases; however, no other major clinical characteristics of CHARGE syndrome could be confirmed with certainty (Table 3). Trio-based CES tests were performed prenatally after genetic counseling, except for case 2, where it was performed postnatally due to the unavailability of stored fetal DNA and maternal refusal to proceed with another invasive procedure during pregnancy. Except for abnormal OS, there were no other clear indications on ultrasound for trio-based CES since CMA was normal and the associated abnormalities were only minor. A de novo heterozygous pathogenic variant in CHD7 was found in cases 1, 2, and 4. All phenotypic features were attributed to this mutation except for case 1, where at least some phenotypic features were possibly related to the superimposed CMV infection. In case 3, all potential pathogenic variants that could be the cause of CHARGE syndrome were ruled out in the trio-based CES. Thus, there was no evidence of a genetic explanation for its clinical presentation. Along with 4 cases exhibiting OS anomalies, CES was performed in 3 additional cases from the study group as part of the genetic investigation of CHD. Among these 3 additional cases, one identified a pathogenic variant in the FILAMIN gene. Finally, during the postnatal follow-up of at least 1 year, no additional diagnosis of CHARGE syndrome was made among CHD cases classified as having normal OS during pregnancy.

In the present study, we demonstrated the feasibility and utility of fetal OS assessment in the diagnosis of CHARGE syndrome in fetuses with major CHDs. These results are aligned with the study of Acanfora et al. [13], which suggested a possible role for ultrasound evaluation of OS in the prenatal diagnosis of CHARGE syndrome. Furthermore, in the current study, abnormal OS were detected in all three cases of CHARGE syndrome, reinforcing previous studies that found arhinencephaly to be a major characteristic of fetal CHARGE syndrome and one of the most prevalent and crucial features of CHARGE syndrome in postnatal MRI and fetal autopsy [11, 18, 19].

Previous reports showed that some of the classical features of CHARGE syndrome do not exist, appear late in pregnancy, or are hard to demonstrate by ultrasound during fetal life due to the need for complex 3D ultrasound techniques, experienced operators, and adequate conditions [11, 18, 20]. In addition, as shown in the current study, CHARGE syndrome also combines multiple and highly variable nonspecific congenital anomalies [6, 11, 18‒21]. Given these challenges, the in utero diagnosis of CHARGE syndrome could be challenging and hard to achieve, highlighting the significant added value of OS assessment in cases of CHDs. Moreover, in the present study, OS evaluation was possible in the vast majority of cases (97%), similar to previous studies, which showed that OS evaluation by ultrasound is simple and feasible [13, 22]. Compared to other major characteristics of CHARGE syndrome, such as facial dysmorphism, coloboma, and ear malformations, the simplicity and good reproducibility of OS assessment could greatly benefit clinical practice, facilitating diagnosis and decision-making in a disease associated with such high morbidity as CHARGE syndrome [8, 20]. Developing OS can be detected by ultrasound starting from 24 weeks of gestation and are seen as fully developed from 28 weeks onward [13, 22]. In the present study, following the guideline recommendation for detailed brain evaluation in cases of fetal CHDs [17], NSG was performed after 30 weeks, a time in which OS and general cortical development are more pronounced and could be assessed more accurately. Although alterations in cortical maturation were previously described in non-syndromic fetal growth restriction and in fetuses that were SGA [23, 24], the pattern of cortical development alterations in these cases is more generalized and usually does not affect only one region or sulcus. Thus, we assume that in case 3, which had no pathogenic gene variations, the late appearance of underdeveloped OS on the day of delivery at 35+3 weeks could not be solely attributed to being SGA.

OS abnormality is also not specific and can be observed in other entities such as Kallmann syndrome and the holoprosencephaly spectrum [14]. However, in the present study, we demonstrated that when combined with major CHD, OS abnormalities have a high positive predictive value for CHARGE syndrome diagnosis. This is extremely important since, occasionally, CHD might be the only significant abnormality when scanning fetuses with CHARGE syndrome [18, 20].

The association between the different types of CHDs and CHARGE syndrome has been extensively documented [11, 20, 25]. In most studies, septal and conotruncal defects are reported as the most frequent CHDs associated with CHARGE syndrome [11, 18, 25]. However, while these anomalies are frequently highlighted in the literature, the significance of wide perimembranous ventricular septal defect in relation to CHARGE syndrome warrants further attention, as this was identified prenatally in all 3 cases of CHARGE syndrome in our study. It is also likely that some cases of CHARGE syndrome with severe conotruncal anomalies opted for early termination of pregnancy before reaching the gestational age at which OS evaluation by NSG could be performed and, therefore, were not diagnosed with CHARGE syndrome.

CHD7 (OMIM #214800) was discovered in 2004 as the primary gene involved in CHARGE syndrome [7]. Although CHD7 mutation is present in most affected individuals, no pathological variant can be found in 5–30% of cases with a clinical diagnosis of CHARGE syndrome [26]. Even though some cases of complete or partial deletion of CHD7 could be diagnosed by a high-resolution CMA [7], single nucleotides or small deletions are by far the most frequent cause of CHARGE syndrome and cannot be detected by this method [7]. Hence, it is not surprising that none of the CHARGE syndrome cases presented in this study were detected by CMA. De novo single nucleotide variants and other rare single-gene syndromes are detectable by CES; however, it is essential to keep in mind that there are also some potential causes for escaped detection by standard CES, gene panel, or single gene. This includes pathogenic variants in noncoding regions (intronic pathogenic variants), particularly in large genes such as CHD7 (188 kb) [26, 27], differences in DNA methylation, cryptic chromosomal rearrangements, and somatic mosaicism [26], which can all lead to a missed diagnosis of CHARGE syndrome. This is especially relevant in cases with subtle phenotypic features, such as in case 3. Whole genome sequencing (WGS) offers several advantages in this context, providing a comprehensive analysis of the genome that enables the detection of pathogenic variants in noncoding regions, structural rearrangements, and other variations that may not be identified by CES. This enhanced capability is especially advantageous for large genes and complex syndromes, making WGS a more effective tool for conditions such as CHARGE syndrome, where noncoding mutations and structural variants can play a significant role. However, WGS is generally more expensive than CES because it generates a larger volume of data and necessitates greater computational resources and specialized expertise, which currently limit its availability in clinical settings.

Although postnatal and autopsy findings of the three cases with pathogenic CHD7 variants in this study did not meet the strict clinical criteria for CHARGE syndrome established by Verloes et al. [9] in 2005, it was later suggested by Hale et al. [28] to combine clinical and molecular diagnosis in the era after molecular testing for CHD7 variants in CHARGE syndrome (i.e., after 2004). Thus, clinicians should not rely solely on the presence or absence of specific features to establish the diagnosis. Updated clinical diagnostic criteria for CHARGE syndrome account for milder phenotypes, and they include pathogenic CHD7 variant status and OS anomalies as major features that are sufficient for diagnosing CHARGE syndrome [11, 28, 29].

From a clinical perspective, the evaluation of OS provides essential information for comprehensive prenatal phenotyping and significantly contributes to interpreting genetic variants detected by ES [2]. Since ES is still not a part of the routine evaluation of fetuses with CHDs, prenatal OS assessment in the standard clinical practice of fetuses with CHD could be greatly useful in selecting those patients with the highest suspicion of CHARGE syndrome to perform ES. This may be particularly important in places with fewer resources where choosing the most suitable cases for ES is essential.

This study has several limitations and considerations to discuss. First, our sample size was modest, and further extensive studies are required to accurately document OS abnormality in CHDs. Second, CES was not performed in cases with normal OS, which could theoretically cause an underdiagnosis of CHARGE syndrome in this group. However, since in these cases, there were no other characteristic anomalies and no later-stage diagnosis of CHARGE syndrome, this probably has no impact on our results. Among the strengths of this study is that it is the first to assess OS in fetuses with CHD by NSG, presenting a complete characterization of this important fetal population. We also highlight the excellent specificity and negative predictive value for OS assessment in the group of fetuses with major CHDs, as no other cases of CHARGE syndrome were diagnosed postnatally in those fetuses classified as having normal OS. Moreover, using a relatively simple methodology, we demonstrated an excellent proportion of agreement between two independent, trained, and experienced neurosonographers (100% agreement). This undoubtedly supports the implementation of OS assessment in clinical work.

In conclusion, a third-trimester systematic evaluation of OS in CHD fetuses is feasible and could improve the detection rates of CHARGE syndrome. Thus, based on this study’s results, we strongly support the routine OS assessment when examining fetuses with major CHDs. These results may be relevant for future studies focusing on predicting and diagnosing prenatal CHARGE syndrome.

This study protocol was reviewed and approved by the Institutional Ethics Committee of BCNatal – Fetal Medicine Research Center, approval reference HCB/2015/0365, and all patients provided written informed consent.

The authors have no conflicts of interest to declare. All coauthors have seen and agreed with the contents of the manuscript, and there is no financial interest to report. We certify that the submission is original work and is not under review at any other publication.

The research leading to these results has received funding from Cerebra Foundation for the Brain Injured Child (Carmarthen, Wales, UK) and ASISA Foundation. The funder had no role in the design, data collection, data analysis, and reporting of this study.

R.C. have made contribution to the design of the article, analysis and interpretation of data, and to the writing of the manuscript. M.P.-C. collected the data and have made substantial contribution to the design of the article and analysis and interpretation of data for the article. M.P.-C. also drafted the article and approved the version to be published. N.M., M.I., E.M., J.M.M-C., and O.G. collected the data and drafted the article. A.B., M.G.-C., M.R.-P., and M.B. collected the data. E.E. collected the data, have made substantial contribution to the design of the article and analysis and interpretation of data for the article, and also drafted the article and approved the version to be published.

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.