Prenatal Microarray Incidentally Diagnoses a Novel Inherited NF1 Intragenic Duplication

ABSTRACT

Introduction

Neurofibromatosis Type 1 (NF1) is a well‐described autosomal dominant genetic condition with variable expressivity. Most commonly caused by intragenic heterozygous loss of function variants or whole gene deletions in the NF1 gene, there are few documented cases of NF1 due to partial gene duplication.

Case Summary

Here we present the case of a proband with an undisclosed clinical diagnosis of NF1 who had been lost to follow‐up as a teenager. She subsequently received a molecular diagnosis of NF1 following prenatal chromosome microarray. Amniocentesis was performed due to a high probability for trisomy 21 on first trimester combined screening and incidentally detected a novel 17q11.2 duplication. RNA analysis confirmed the pathogenicity of this partial gene duplication involving the NF1 gene, but not before the proband conceived a second pregnancy which also carried the 17q11.2 duplication. This resulted in a mother and baby with differing clinical features of NF1. The mother presented in childhood with the more common features of café‐au‐lait macules, bilateral axillary freckling, and later left forearm subcutaneous lipoma, while her son presented in the neonatal period with the less common isolated tibial pseudoarthrosis. Therefore, this demonstrates intrafamilial phenotypic variability.

Contribution

This case describes a pathogenic 17q11.2 duplication not previously reported in NF1 literature as well as demonstrating the ongoing importance of referral to a clinical genetics service in the event of any unexpected result on prenatal microarray, and the importance of thorough medical and family history elicitation, pre‐test counseling, and informed consent.

Incidental molecular diagnosis of Neurofibromatosis Type 1 when prenatal chromosome microarray detected a duplication encompassing exons 2–36 of the NF1 gene. Pathogenicity was confirmed by RNA analysis. This case identifies a novel pathogenic NF1 duplication and highlights the importance of clinical genetics referral, pre‐test counseling, and informed consent.

1. Introduction

Neurofibromatosis Type 1 (NF1) is an autosomal dominant, inherited genetic condition with a global prevalence of 1 in 3164 (Lee et al. 2023). The most commonly seen features of NF1 include multiple café‐au‐lait macules (CALMs), axillary freckling, neurofibromas, lisch nodules, optic nerve glioma, and osseous lesions such as pseudarthrosis (Castle et al. 2003; Legius et al. 2021; Van Minkelen et al. 2014). A number of clinical features of NF1 represent an increased risk of malignancy, with regular lifetime surveillance advised for individuals with a clinical or molecular diagnosis of NF1 (Stewart et al. 2018; Evans et al. 2017). Typically, patients with NF1 will initially receive a diagnosis based on clinical criteria (Legius et al. 2021), and with advancements in genomic testing it has become commonplace for subsequent confirmation via molecular testing (Griffiths et al. 2007). Clinical diagnostic criteria for NF1 were revised in 2021 to include updated clinical and genetic understandings, as well as better differentiation between NF1 and Legius syndrome (Legius et al. 2021). Seven clinical and molecular features are included in the diagnostic criteria for an individual with an unaffected parent. The presence of two or more features are sufficient for a diagnosis of NF1. These include but are not limited to; six or more CALMS, axillary or inguinal freckling, an osseous lesion such as a pseudarthrosis, and a germline pathogenic NF1 variant (Legius et al. 2021). When a parent has a diagnosis of NF1, their child meets diagnostic criteria if they have one or more of the clinical or molecular features (Legius et al. 2021).

Although few case reports of antenatal detection of NF1 features via prenatal ultrasound or magnetic resonance imaging exist, this is uncommon and typically involves non‐specific findings or advanced tumor detection in the third trimester (McEwing et al. 2006; Yang et al. 2022). Without reliable sonographic indicators for identifying NF1 increased risk, detailed parental health and family history become the most important information to identify such cases. In an Australian setting, the first opportunity for this would likely be when attending an antenatal clinic during the first trimester of pregnancy. However, the volume of information required to be collected by a midwife to assess fetal and maternal risk factors can take up to 52 min per patient, and even longer if assistance is required after identifying an increased risk (Waller et al. 2018). With limited consultation time and/or practitioner knowledge, family health history is often underutilized in many medical settings (Ginsburg et al. 2019).

Heterozygous intragenic variants involving the neurofibromin 1 (NF1) gene (NM_000267.3; OMIM #613113) account for the majority of cases of NF1, with only 5%–13% of molecular diagnoses identifying a whole or partial gene deletion (Kluwe et al. 2004; Griffiths et al. 2007; Bianchessi et al. 2015). Reports of whole or partial gene duplications causing NF1 are limited (Bianchessi et al. 2015; Giugliano et al. 2019). The identification and publication of new NF1 pathogenic variants will add to the existing literature of variants known to be associated with NF1 disease. This may also assist in supporting a molecular diagnosis for those with previously uninformative testing.

The NF1 gene is a tumor suppressor gene comprising up to 61 exons located at 17q11.2 (DeClue et al. 1992; Griffiths et al. 2007; Trovo‐Marqui and Tajara 2006). The resulting neurofibromin protein regulates multiple intracellular processes, including negative regulation of the RAS‐cyclic AMP and RAS‐MAPK pathways (Ratner and Miller 2015; Trovo‐Marqui and Tajara 2006). Expression of neurofibromin mostly occurs in neurons, Schwann cells, oligodendrocytes, and leukocytes, with evidence of reduced expression in certain tumors for patients with or without NF1 (Johnson et al. 1993). Caused by loss of function variants, the clinical features of NF1 show variable expressivity between and within families, with limited established genotype–phenotype correlations (Castle et al. 2003; Van Minkelen et al. 2014).

Prenatal genetic counseling in the context of NF1 is complicated by the variable severity and expressivity, lifelong burden of disease, and potential for intellectual disability (Grisart et al. 2008; Stewart et al. 2018; Radtke et al. 2020). Following the Abortion Law Reform Act in 2019, the state of New South Wales, Australia, allows for termination of pregnancy up to 22 weeks gestation without restriction (NSW Health 2021). This provides couples with greater reproductive and decision‐making options, especially in cases when one partner is known to have a pathogenic variant in the NF1 gene. However, many couples express difficulty in electing for termination of pregnancy for NF1 due to the variable disease severity; therefore, they decline prenatal testing and instead opt for in vitro fertilization (IVF) and pre‐implantation genetic testing of embryos where feasible (Crawford et al. 2015).

Here we describe a case of a novel 17q11.2 duplication involving exons 2–36 of the NF1 gene (NM_000267.3), which was detected incidentally on prenatal microarray. The clinical indication for prenatal diagnostic testing was a high probability of trisomy 21 on first trimester aneuploid screening. Phenotyping and subsequent RNA analysis demonstrated a tandem duplication and supported reclassification of the variant and a diagnosis of NF1 for the mother of the fetus. The mother later declined prenatal diagnostic testing for NF1 in a subsequent pregnancy.

2. Case Summary

In addition to the identification of a novel intragenic NF1 pathogenic variant, this case highlights the importance of thorough medical and family history collection in the antenatal setting. Attention should be paid to possible heritable conditions which would warrant prenatal genomic testing. Proceeding with prenatal genomic testing for any reason without this exploration may also pose a risk of unintended findings.

The proband has consented to publication of her case. The proband was a nulliparous 39‐year‐old female referred for genetic counseling following elective termination of pregnancy at approximately 16 weeks' gestation following amniocentesis and prenatal diagnosis of trisomy 21. The pregnancy was conceived via IVF with a history of infertility and low sperm motility for her reproductive partner. Preimplantation embryo screening for chromosome aneuploidy was not performed, and the couple had both returned normal karyotype results as part of the work‐up for IVF. Amniocentesis was performed due to a high‐probability result for trisomy 21 on combined first trimester screening (nuchal translucency ultrasound and maternal serum screening). Termination of pregnancy was elected following receipt of the rapid aneuploidy result which confirmed a diagnosis of trisomy 21 on DNA extracted from amniocytes. The Royal Australian and New Zealand College of Obstetricians and Gynaecologists advise that rapid aneuploidy screening be used as an adjunct only to prenatal chromosome microarray (CMA) (RANZCOG 2018). Subsequently, an 850K CytoSNP v1.3 SNP CMA with a mean resolution of 0.05 Mb was performed on DNA from extracted amniocytes and reported a 0.131 Mb 17q11.2 duplication: arr[GRCh38] 17q11.2(31155363_31286582)x3, in addition to the trisomy 21. The identified duplication encompasses a region of the NF1 genomic locus, with breakpoints smaller than those reported for the established recurrent NF1 microduplication syndrome. The variant is absent from population and disease‐associated databases with current evidence insufficient to support triplosensitivity of the duplicated region or an effect on NF1 function. Accordingly, as per the criteria 1A, 2I, 3A, the duplication was classified as a Class 3 variant of uncertain significance (VUS), given its uncertain potential to disrupt NF1 gene function.

The proband (II‐2) is the third of four children born to an unrelated couple of Chilean ancestry (see Figure 1). Her mother was diagnosed with mild scoliosis later in life, but the family history is otherwise unremarkable.

FIGURE 1.

Pedigree.

At the proband's first contact with the genetics team, a telephone intake consultation with a genetic counselor revealed a childhood clinical diagnosis of neurofibromatosis 1 that had not been reviewed since her early teens. She had not disclosed this information to any of her current health‐care providers. With consent, historical medical records were obtained which confirmed that the possibility of NF1 was first raised at 2 years old due to increasing number and size of CALMs. In 1991 at 8 years old, the proband satisfied the diagnostic criteria of NF1. Now at 39 years old, review of medical records and examination with a clinical geneticist identified the following NF1 diagnostic criteria (Legius et al. 2021) and typical phenotypic features: > 6 CALMs, bilateral axillary freckling, left forearm subcutaneous lipoma, macrocephaly and NF1‐related interstitial lung disease. The proband's CALMs and bilateral axillary freckling alone are sufficient to meet NF1 diagnostic criteria.

Targeted parental CMA performed on DNA extracted from peripheral blood confirmed that the 17q11.2 duplication detected prenatally was maternally inherited. Testing was again performed using the 850K CytoSNP v1.3 SNP microarray platform with a mean resolution of 0.05 Mb. See Figure 2 for a schematic representation of the NF1 gene and intragenic duplication.

FIGURE 2.

Schematic representation of the genomic structure of the NF1 gene (transcript NM_000267.3). Each exon is represented by a yellow box and labeled with its corresponding exon number. The blue arrow indicates that the exons continue sequentially without interruption and are omitted from the schematic for clarity. Red boxes highlight the exons encompassed by the 17q11.2 duplication identified in our patient, illustrating their location within the NF1 gene. NF1 exon numbering is according to the LRG214t1.

DNA from peripheral blood was sent to UAB Medicine in Birmingham, AL, USA for next generation sequencing of the NF1 and SPRED1 genes. Pathogenic variants in the SPRED1 gene are associated with a NF1‐like phenotype known as Legius syndrome, which does not include the presence of neurofibromas or Lisch nodules (Brems et al. 2007; Legius et al. 2021). As is routine practice for the clinical genetics team arranging diagnostic testing for suspected NF1, sequencing of the SPRED1 gene was also requested to exclude this differential diagnosis. A VUS c.(60+599_93)_(4772+3_4772+30133)dup was reported in the NF1 gene (NM_000267.3). No reportable variants were detected in SPRED1.

An additional peripheral blood sample was then sent to UAB Medicine to perform RNA analysis. This involved lymphocyte culture, RNA and DNA extraction, PCR and RT‐PCR, gDNA‐ and RNA‐based Sanger sequencing of NF1 exon 2 and 36. At this time, the proband was 17 weeks pregnant with her second IVF conceived pregnancy. She requested that the result of RNA analysis not be disclosed until after the birth of her baby due to her increasing anxiety about the potential implications should the NF1 intragenic duplication be reclassified as pathogenic. RNA analysis demonstrated exon 36 linking to exon 2, confirming that the duplication was in tandem and is expected to be an out‐of‐frame duplication resulting in nonsense‐mediated mRNA decay. Additionally, the size of the duplication, at almost one third of the total gene size, is suspected to be very disruptive. Here, the results of RNA analysis provided additional support for the pathogenicity of this 17q11.2 intragenic duplication. The duplication was reclassified as Class 5, pathogenic in accordance with ACMG variant interpretation recommendations (Richards et al. 2015). Therefore, confirming a molecular diagnosis of NF1 for the proband, and a 50% chance of transmission to any offspring she conceived. As per her wishes, the treating team did not contact the proband with these results, instead planned to attempt to re‐engage with her at 3–6 months postpartum.

Unbeknownst to the clinical genetics team, the proband's newborn son presented to hospital at 6‐weeks‐old for management of a viral upper respiratory tract infection. During this hospital presentation, a lump was noted on the newborn's leg which was later confirmed on imaging to be tibial pseudoarthrosis. Tibial pseudoarthrosis can be seen in individuals with NF1 and is included in the clinical diagnostic criteria (Legius et al. 2021). A CMA arranged by the Pediatrician confirmed that the newborn had inherited the maternal 17q11.2 duplication. Although he did not have CALMs or other identifiable manifestations of NF1 at this time, as the child of a parent with NF1, the presence of one or more diagnostic criteria in the newborn was sufficient for his own diagnosis of NF1 (Legius et al. 2021). The marked intrafamilial variability is consistent with the well‐recognized variable expressivity of NF1 and suggests that additional genetic, epigenetic, or environmental modifiers contribute to phenotypic severity beyond the primary pathogenic variant. The detection of the maternally inherited 17q11.2 duplication via CMA precludes the need for more costly and time‐consuming molecular testing which may not be available or financially possible in certain cases. At the public hospital service where the proband and her son were seen, public health funding covers the cost of genetic testing for eligible patients.

3. Discussion

Targeted RNA analysis of the NF1 gene c.(60+599_93)_(4772+3_4772+30133)dup identified in the proband demonstrated that the intragenic duplication occurred in tandem, involving exons 2–36. Large intragenic duplications, as identified in this family, are predicted to disrupt normal gene function through abnormal splicing or frameshift‐mediated nonsense‐mediated mRNA decay, ultimately resulting in loss of neurofibromin function. However, the clinical consequences of these variants appear to be variable. Whole‐gene NF1 deletions result in haploinsufficiency due to loss of one functional copy of the gene and are associated with a more severe clinical phenotype, including an increased burden of neurofibromas, cognitive impairment, and dysmorphic features (Mensink et al. 2006; Van Minkelen et al. 2014; Giugliano et al. 2019). In contrast, the pathogenic mechanisms and genotype–phenotype correlations associated with intragenic NF1 duplications remain poorly characterized.

The majority of NF1 cases can be attributed to point mutations or single nucleotide variants in the NF1 gene (Bianchessi et al. 2015; Kang et al. 2020), with only up to 11% demonstrating copy number loss or microdeletions involving the entire NF1 gene, with or without loss of flanking genes (Kluwe et al. 2004; Zhang et al. 2015). Reported instances of pathogenic copy number gains or micro/intragenic duplications of the NF1 gene are rare (Kehrer‐Sawatzki et al. 2014; Bianchessi et al. 2015; Giugliano et al. 2019). Considering all pathogenic variants involving the NF1 gene, approximately 2% are described as intragenic copy number variants (Messiaen and Wimmer 2008; Wimmer et al. 2006). In a study of 281 individuals with a clinical diagnosis of NF1, Giugliano et al. (2019) were able to confirm a pathogenic variant in the NF1 gene for 206 probands (73.3%). Only one case (0.5%) was found to have an intragenic duplication which involved exons 37–51 (Giugliano et al. 2019). This intragenic duplication had not been reported previously. The female patient presented with > 6 CALMs and a neurofibroma at 13 years old, and by 15 years had developed freckling (Giugliano et al. 2019). In a similar cohort of 293 unrelated probands, Bianchessi et al. (2015) confirmed novel intragenic duplications of NF1 in two of 206 molecular diagnoses (< 1%). Phenotype information for these individuals is not available.

In this case, the molecular diagnosis of NF1 in an adult came about as an unintended consequence of prenatal diagnostic testing for trisomy 21. With an early diagnosis of NF1 for the proband's son, he now has the opportunity for early monitoring for potentially medically or developmentally significant NF1‐related clinical features. Whilst the proband's son satisfies clinical diagnostic criteria for NF1, his presentation is less commonly seen in individuals with NF1. For those diagnosed with congenital tibial pseudoarthrosis (CTP), studies have found that 55%–84% will be diagnosed with NF1, and the remainder pose an etiological mystery (Kjell et al. 2016; Zhu et al. 2019). In a study of 245 probands (mean age 14 ± 12) suspected to have NF1, 150 met clinical diagnostic criteria for NF1, and of those, 99.3% presented with CALMs and only 2% presented with a bone dysplasia such as CTP (Giugliano et al. 2019). Similar figures have been reported in other publications (Jett and Friedman 2010; Van Minkelen et al. 2014). Without his mother's diagnosis and unintended prenatal finding, this newborn may have had a much longer diagnostic journey.

Prenatal CMA in the presence of advanced maternal age, positive family history, or screening abnormality will identify a VUS in approximately 4%–6% of cases (Mardy et al. 2021; Mitrakos et al. 2023). However, depending on local guidelines, some laboratories only report a VUS if it is in a gene associated with the described fetal phenotype (Vears et al. 2017; Lewis et al. 2021), so the true incidence may be greater. An alternative laboratory may not have reported the prenatal 17q11.2 duplication; therefore, missing an important diagnosis and point of re‐engagement for this mother and her future children. With an increasing number of prenatal CMAs being requested by non‐genetics professionals (Bajguz et al. 2022), this case also demonstrates the ongoing importance of referral to a clinical genetics service in the event of any unexpected result on prenatal microarray as well as the importance of thorough medical and family history elicitation, pre‐test counseling, and informed consent.

Author Contributions

Rhiannon Sandow: writing – original draft, resources. Gunjan Garg: supervision, writing – review and editing, conceptualization, investigation.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this case are available from the corresponding author upon reasonable request.