Abstract
Introduction: Brittle cornea syndrome (BCS) is a rare connective tissue disorder with ocular and systemic features. Extreme corneal thinning and fragility are the main hallmarks of BCS. Case Report: A 4-year-old boy presented with recurrent spontaneous corneal perforation. He had blue sclera, corneal leucoma, irregular iris, shallow anterior chamber, corneal astigmatism, and bilateral corneal thinning. He also had several systemic features including hearing loss, skin hyperelasticity, joint hypermobility, scoliosis, and umbilical hernia. A diagnosis of BCS was confirmed with molecular analysis. A homozygous c.17T>G, p.(Val6Gly) variation was identified in the PRDM5 gene. Discussion: p.(Val6Gly) variation in PRDM5 was previously reported in 2 patients with BCS. We also considered PRDM5 c.17T>G, p.(Val6Gly) variation as pathogenic based on the following features: the absence of the variation in population databases, in silico predictions, segregation analysis, and clinical signs of our patient. Extremely thin and brittle corneas lead to corneal perforation spontaneously or after minor trauma. Nearly all patients have lost their vision because of corneal rupture and scars. The key challenge in the management of BCS is the prevention of ocular rupture which relies on early diagnosis. Early diagnosis allows for taking prompt measures to prevent ocular rupture.
Established Facts
Brittle cornea syndrome (BCS) is a rare hereditary connective tissue disorder that results from pathogenic variants in either the ZNF469 (MIM 612078) or PRDM5 (MIM 614161) gene.
The c.17T>G, p.(Val6Gly) variant in the PRDM5 gene was only described in 2 previous cases.
Novel Insights
Our report describes a case with homozygous PRDM5 c.17T>G, p.(Val6Gly) variant, with the typical ocular features of BCS as well as a variety of extraocular manifestations.
Both the previously reported cases and our case with the PRDM5 c.17T>G, p.(Val6Gly) variant are Turkish in origin. Although more cases are needed, considering the rarity of this condition, this variant can be argued as a “founder mutation” for Turkish ethnicity with our case report.
Introduction
Brittle cornea syndrome (BCS; OMIM 229200 and 614170) is a rare connective tissue disorder characterized by extreme corneal thinning and fragility. Other typical ocular findings are blue sclera, keratoglobus, keratoconus, high myopia, and irregular corneal astigmatism [Ticho et al., 1980; Burkitt Wright et al., 2013; Lechner et al., 2014]. Due to the extremely thin corneas (central corneal thickness: 220–450 μm; normal range: 520–560 μm), BCS corneas are unable to maintain their shape and structure under normal biomechanical stresses and are prone to spontaneous rupture. Nearly all patients have lost their vision because of corneal rupture and scars [Royce et al., 1990; Ramappa et al., 2014; Avgitidou et al., 2015]. BCS is a multisystem connective tissue disorder that shares a variety of extraocular manifestations, including auditory, skin, and musculoskeletal features [Khan et al., 2012]. Hearing loss, joint hypermobility, developmental hip dysplasia, muscle hypotonia, scoliosis, skin hyperextensibility, and deformities of the hands and feet are common in BCS [Dhooge et al., 2021].
Since BCS shares general connective tissue fragility, it shows significant overlap with the Ehlers-Danlos syndrome (EDS) spectrum [Dhooge et al., 2021]. BCS has been considered as a phenotypic variant of the kyphoscoliotic form of EDS, previously designated EDS type VIA (MIM 225400), which is caused by biallelic variations in the PLOD1 gene (MIM 153454) encoding lysyl hydroxylase. With the identification of the molecular basis of many forms of EDS and clinical reclassification, it has been clarified that BCS is a separate entity [Walkden et al., 2019]. The genetic etiology of BCS was first identified in 2006 [Dhooge et al., 2021]. Biallelic variations in ZNF469 (MIM 612078) and PRDM5 (MIM 614161) have been identified as being responsible for BCS [Abu et al., 2008; Christensen et al., 2010]. ZNF469 plays a role in normal anterior segment and corneal development and acts as an important determinant of corneal thickness. ZNF469 has also been implicated in the development of keratoconus. PRDM5 is critical for the development and maintenance of the extracellular matrix and strongly associated with corneal thickness [Davidson et al., 2015; Walkden et al., 2019]. It also affects the development of retinal microvasculature and Bruch’s membrane [Walkden et al., 2019; Selina et al., 2020]. It was shown that both ZNF469 and PRDM5 variations disrupt collagen deposition and collagen fibril assembly [Walkden et al., 2019].
This report describes a boy with BCS who had recurrent corneal perforation.
Case Report
A 4-year-old boy was referred to the pediatric genetics department because of recurrent corneal perforation and vision loss. He was born in the 36th week of gestation with normal spontaneous vaginal delivery. His birth weight was 1,600 g. Pregnancy was complicated by intrauterine growth restriction. He was hospitalized for a month in the neonatal care service because of prematurity and low birth weight. He was the second child of a healthy fourth-degree consanguineous couple (Fig. 1). Family history was negative for known genetic disorders; to the best of the parents’ knowledge, there is nobody in the family with an ocular phenotype like their son. He had an operation for developmental dysplasia of the hip when he was 1 year old. He walked independently at the age of 2 years and formed sentences. When he was 3 years old, he had a spontaneous corneal rupture of the right eye for the first time. There were no ocular traumas according to the parents, but a corneal rupture recurred in both eyes 6 months ago. He had a history of easy bruising. The child had a specialized education program and underwent physiotherapy for a year. His weight was 14 kg (10th percentile), height was 103 cm (25th–50th percentile), and his head circumference was 51 cm (25th–50th percentile). His physical examination was distinctive, with a wide forehead, blue sclera, corneal leucoma, strabismus, prominent eyeballs, sparse teeth with caries, pectus excavatum, generalized joint laxity (with a Beighton mobility score of 7), umbilical hernia, scoliosis, mild muscular hypotonia, soft and flexible skin, and ecchymosis on the extremities (Fig. 2). Echocardiogram was normal. The hearing assessment revealed mild mixed-type hearing loss in the right ear and moderate deafness for higher frequencies in the left ear. Ophthalmological examination revealed bilateral corneal leucoma secondary to previous corneal perforations, blue sclera, an irregular iris, shallow anterior chamber, keratoconus, and corneal astigmatism. He had no retinal abnormalities or retinal detachment. The central corneal thickness measured 300 μm in the left eye and 250 μm in the right eye. Medical history, physical examination, and ocular findings prompted a preliminary diagnosis of BCS or kyphoscoliotic form of EDS. To differentiate these 2 conditions, lysyl hydroxylase activity was examined by urine, and deoxypyridinoline/pyridinoline ratio was found to be normal in our patient. Genetic analysis was requested for heritable connective tissue disorders including BCS and a kyphoscoliotic form of EDS.
Genomic DNA was isolated from nucleated peripheral blood cells by using the EZ1 system (Qiagen, Hilden, Germany). Libraries were prepared with the QiaSeq Custom Amplicon Panel kit according to the manufacturer’s instructions. Genes included in this panel were ADAMTS2, ATP7A, B3GALT6, B3GAT3, B4GALT7, CHST14, COL12A1, COL1A1, COL1A2, COL3A1, COL5A1, COL5A2, DSE, EFEMP2, ELN, FBLN5, FBNFKBP14, FLNA, GORAB, LTBP4, PLOD1, PRDM5, PYCR1, RIN2, SLC39A13, TNXB, and ZNF469.
Amplified libraries were sequenced on the Illumina MiSeQ system. Ingenuity Variant Analysis Software (Qiagen) was used to analyze the data, and the IGV_2.3.8 program was used to evaluate the data visually.
In the patient, sequencing revealed a homozygous variant: NM_018699.3 (PRDM5):c.17T>G, p.(Val6Gly). This was a variant of unknown significance (PM2, PP3, PP4, PP5, BP1) according to ACMG 2015 classification guidelines. The variant was submitted to ClinVar (RCV001260235) and predicted as likely pathogenic. In silico analysis revealed the following: Mutation Taster, disease-causing; SIFT, damaging; DANN score, 0.9781; and Gerp score, 3.5699. The variant NM_018699.3 (PRDM5):c.17T>G, p.(Val6Gly) was not reported in the ExAc or gnomAD population databases and has been reported in the HGMD® Professional 2019.3 database as a disease-causing variation, with the accession number CM157834. The parents are both heterozygous for the same variation. Screening for the rest of the family was planned.
Discussion
BCS is a rare autosomal recessive connective tissue disorder. The prevalence is estimated to be less than 1/1,000,000 [Walkden et al., 2019]. It was first described by Stein et al. [1968] in 2 siblings of Tunisian Jewish descent. Since then, additional cases have been reported worldwide [Walkden et al., 2019], resulting in a total of 86 patients [Selina et al., 2020; Dhooge et al., 2021].
BCS is a connective tissue disorder with ocular and systemic features. Typical clinical features of BCS are summarised in Table 1. The ophthalmic features are the most devastating. Extremely thin and brittle corneas lead to corneal perforation spontaneously or after minor trauma. Ocular perforation was reported in more than two-thirds of patients by Dhooge et al. [2021]. Corneal perforation is frequently observed at a mean age of 4.3 years (range 1.5–19) [Walkden et al., 2019]. Like corneal perforation, keratoconus, keratoglobus, and severe myopia can cause vision loss in BCS patients. More than half of the reported cases were complicated by permanent vision loss [Walkden et al., 2019]. Blue sclera is the most common ocular finding of BCS. It was noted in 72 of 78 patients by Dhooge et al. [2021]. Blue sclera is correlated with central corneal thickness reduction, but is not specific to BCS and can be seen in other conditions like EDS and osteogenesis imperfecta (OI). Retinal detachment and secondary glaucoma have also been reported [Walkden et al., 2019].
BCS patients generally demonstrate extraocular symptoms. Dhooge et al. [2021] reported that only 8 of the 78 patients had no systemic findings. Hearing impairment was documented in 32 of 78 patients. Deafness can be conductive, sensorineural, or mixed type. Progressive deafness, especially for higher frequencies, along with a hypercompliant tympanic membrane can be seen [Burkitt Wright et al., 2011, 2013]. Thus, audiometry and tympanography should be performed for all BCS cases. Joint hypermobility, especially in small joints, has been noted in the majority of affected patients. Joint hypermobility was reported in 64 of the 78 patients with BCS; in 13 patients it was the only extraocular feature [Dhooge et al., 2021]. Other musculoskeletal findings are congenital hip dysplasia, scoliosis, foot and hand deformities, arachnodactyly, pectus deformity, muscle hypotonia, congenital hip dislocation, contractures, delayed motor development, and bone fragility [Walkden et al., 2019; Dhooge et al., 2021]. Some recent reports suggest that BCS may have bone fragility and fracture phenotype [Burkitt Wright et al., 2013; Cundy et al., 2021; Dhooge et al., 2021]. Therefore, BCS should be kept in mind in patients with fractures and a history of osteopenia as a differential diagnosis. Skin findings of BCS include soft, smooth, hyperextensible, thin, fragile, and transparent skin and easy bruising, Abnormal scarring, if present, is usually mild [Burkitt Wright et al., 2013; Avgiditidou et al., 2015; Wan et al., 2018]. Hernias, dental problems, and cardiac defects are less frequently seen in BCS. Facial dysmorphic features were defined in some reports as frontal bossing, a high-arched palate, and a depressed nasal bridge [Dhooge et al., 2021].
Variations in the ZNF469 (encoding zinc finger protein 469) and PRDM5 (encoding the PR domain-containing protein) genes are causative for BCS1 (MIM 229200) and BCS2 (MIM 614170), respectively. The ZNF469 gene is a single-exon gene located at 16q24 and encodes 3,953 amino acid residues. PRDM5 is a 16-exon gene that has a PR SET domain and C2H2 zinc finger domains located at 4q25q26; it encodes 630 amino acid residues [Porter et al., 2015; Wan et al., 2018]. Both genes encode proteins with multiple zinc fingers, suggesting roles in transcription. The PRDM5 gene product acts as a transcriptional regulator and participates in pathways regulating the extracellular matrix (ECM). Besides the regulatory role in ECM development, it was shown that PRDM5 has a direct role in the regulation of collagen genes. To date, 53 BCS1 patients with ZNF469 variations and 33 BCS2 patients with PRDM5 variations have been reported [Selina et al., 2020; Dhooge et al., 2021]. In total, 24 different variations in the ZNF469 gene (18 homozygous, 6 compound heterozygous) and 14 different variations (all homozygous) in the PRDM5 gene were identified [Dhooge et al., 2021]. Although the exact mechanisms of ZNF469 and PRDM5 genes in BCS are not understood yet, the high proportion of frameshift and nonsense variations among the reported variations of both genes suggests that they act as “loss-of-function alleles” [Galli et al., 2012; Burkitt Wright et al., 2013]. The PRDM5 c.17T>G, p.(Val6Gly) missense variation is located within the PR-SET domain of the gene, which is thought to have a role in protein-protein interactions. This variation is not found in gnomAD. It affects a highly conserved amino acid. The majority of the in-silico prediction programs suspect the deleterious effects of amino acid substitution. The variation is found in ClinVar and predicted as likely pathogenic. So, due to the combined information on the clinical phenotype of our patient, segregation analysis, the absence of the variation in population databases, and in silico predictions, we also suspect that PRDM5 c.17T>G, p.(Val6Gly) is causative for BCS. The homozygous PRDM5 c.17T>G, p.(Val6Gly) variant was first described in a 3-year-old boy from Germany whose parents were Turkish in origin. He presented with ocular manifestations only [Avgitidou et al., 2015]. The second patient with this variation was a 34-year-old female with an initial diagnosis of OI [Dhooge et al., 2021]. Along with BCS ocular features, she had hearing loss and joint hypermobility. A comparison of the clinical findings in the previous 2 patients and our case with the same mutation is shown in Table 2 [Avgitidou et al., 2015; Dhooge et al., 2021]. The second patient was also of Turkish origin [Dhooge et al., 2021]. Although more cases are needed, considering the rarity of this condition, this variation can be argued as a “founder mutation” for Turkish ethnicity.
The list of differential diagnoses for BCS is long. Kyphoscoliotic type 1 EDS, caused by biallelic variations in the PLOD1 gene (kEDS-PLOD1) and formerly known as EDS VIA, is at the top of the list. The other conditions are musculocontractural types of EDS (EDSMC; MIM 601776 and 615539), kyphoscoliotic type 2 EDS (kEDS-FKBP14; MIM 614557), spondylodysplastic type EDS (EDSSPD; MIM 130070, 612350, and 615349), OI (MIM 166200), Marfan syndrome (MIM 154700), and corneal endothelial dystrophy (CHED; MIM 217700) [Ramappa et al., 2014; Dhooge et al., 2021]. Clinical differentiation between kEDS-PLOD1 and BCS may be challenging. In kEDS-PLOD1, PLOD1 variations cause lysyl hydroxylase enzyme deficiency. The deficient activity of lysyl hydroxylase leads to an increase in the ratio of deoxypyridinoline/pyridinoline cross-links in urine, which is considered one of the reliable diagnostic markers for kEDS-PLOD1. The deoxypyridinoline/pyridinoline ratio is elevated in kEDS-PLOD1, whereas it is normal in BCS [Ramappa et al., 2014]. BCS presents with a striking ocular phenotype. Ocular rupture is less frequently pronounced in kEDS-PLOD1 than in BCS, and when it occurs, it tends to be scleral rather than corneal. kEDS-PLOD1 individuals have more pronounced generalized connective tissue manifestations. Scoliosis, muscular hypotonia, and vascular complications are more prevalent and severe in kEDS-PLOD1 than in BCS [Rohrbach et al., 2011; Ramappa et al., 2014]. Thus, the early recognition and diagnosis of both conditions are very important for necessary preventive measures to be taken to reduce the rate of morbidity associated with these diseases.
The key challenge in the management of BCS is the prevention of ocular rupture which relies on early diagnosis. Early diagnosis allows prompt measures to be taken to prevent ocular rupture. To protect the ocular surface, special protective glasses should be prescribed [Zlotogora et al., 1990; Walkden et al., 2019]. It is mandatory to train lifestyle behaviors to reduce eye-hand contact and eye rubbing in patients, and also their parents, other caregivers, and school staff. Contact sports should be avoided. Contact lens usage is restricted because of corneal thinning and the risk of trauma. Surgical repair is necessary in cases of corneal rupture [Ramappa et al., 2014; Davidson et al., 2015; Walkden et al., 2019]. There were some therapeutic methods like epikeratoplasty, corneoscleral grafting, and collagen crosslinking, but the efficiency of the treatment is unsatisfactory and limited, with some complications [Vajpayee et al., 2002; Kanellopoulos and Pe, 2005; Walkden et al., 2019].
In conclusion, with this case report, we aimed to highlight the importance of the early recognition of BCS to permit appropriate management. The prevention of ocular rupture is the most important step in the management of BCS. Protective glasses and lifestyle modifications minimize the risk of corneal rupture. Ophthalmological long-term follow-up and general follow-up examinations are needed to prevent ocular and systemic complications of the disease.
Statement of Ethics
Ethical approval was not required for this study by the local ethics committee of Samsun Education and Research Hospital. Written informed consent was obtained from the parents of the patient for publication of the medical condition of their son and images.
Conflict of Interest Statement
The authors declare that there is no conflict of interest regarding the publication of this paper.
Funding Sources
There are no sponsorship or funding arrangements relating to our research.
Author Contributions
Aslıhan Sanrı: writing and revision of the manuscript. Selma Demir and Hakan Gurkan: molecular analysis, review of the manuscript.
Data Availability Statement
The dataset used and analyzed during the current study is available from the corresponding author on reasonable request.