Exome Sequencing as a Potential Diagnostic Adjunct in Sporadic Congenital Hydrocephalus

Abstract

This study evaluates whole-exome sequencing as a diagnostic tool in a cohort of neurosurgically treated congenital hydrocephalus probands.

Congenital hydrocephalus (CH) affects 1 in 1000 births, is a major cause of morbidity, and costs the US health care system $2 billion annually.¹ More than 40% of CH cases are thought to have a genetic etiology. However, only less than 5% of CH cases are associated with a defined gene mutation,¹ limiting the utility of genetic testing with targeted approaches and underscoring the need for CH gene discovery. The X-linked recessive form of hydrocephalus associated with aqueductal stenosis (OMIM #307000) owing to mutation of L1 cell adhesion molecule (L1CAM) is the most common form of CH (approximately 3% of total cases),¹ and variably presents in male patients with mental retardation, spastic paraparesis, and adducted thumbs as part of L1 syndrome.² Currently, L1CAM is the only gene for which targeted screening is advised and routinely clinically available for CH; however, there is little consensus on testing criteria or methods.³ Moreover, patients with no family history who present with perinatal CH are most commonly assessed, if at all, by karyotype or microarray analysis aimed at detecting chromosomal abnormalities or copy number variants but unable to identify de novo or inherited rare mutations.⁴

Methods

We applied whole-exome sequencing (WES) to what is, to our knowledge, the largest cohort of neurosurgically treated CH probands collected to date (475 probands and 284 case-parent trios). The structure of our study, in which patients were ascertained on the basis of not having a clinical syndromic or genetic diagnosis from a multitude of institutions largely via a social media approach, enabled a “real-life” snapshot of CH management and allowed us to examine how frequently L1CAM-mutant hydrocephalus is going undiagnosed. The data were collected and analyzed in 2020. Institutional review board approval was provided by Yale University, and all participants provided written consent.

Results

We uncovered 5 novel and 7 total protein-damaging L1CAM mutations (1.47% of our cohort) in male CH probands (Table). Clinical information was available for 6 of 7 patients. Of these 6, all presented with hydrocephalus and aqueductal stenosis diagnosed prenatally (4 or 6) or shortly after birth (2 of 6) with ultrasonography or magnetic resonance imaging. A total of 2 of 7 patients received genetic screens, one of whom was first assessed via microarray-based comparative genomic hybridization at birth, yielding a negative result. The remaining 5 patients harbored mutations that would not have been detected using traditional chromosomal or microarray-based approaches. Novel L1CAM mutations included the p.Trp460Cys and p.Trp635Arg deleterious missense mutations, mapping to the fifth (of 6) immunoglobulin-like domain and the first (of 5) extracellular fibronectin domain, respectively; 2 splice site mutations, c.1828 + 1G>A (localizing to intron 15) and c.1546 + 1G>T (located in intron 13) were associated with loss-of-function translation frameshifts; and the stop-gain mutation p.Glu304X. The previously reported deleterious missense mutation p.Val788Phe (localizing to the second fibronectin domain),⁵ and the known splice site mutation c.806 + 1G>C (positioned in intron 8),⁶ were also identified (Figure). All 6 patients for whom clinical information was available had neurodevelopmental delays. Classic L1 syndrome findings of bilateral adducted thumbs, hypotonia, cerebral palsy, epilepsy, and white-matter hypoplasia were each seen in 3 of 6 patients. Agenesis of the corpus callosum, macrocephaly, and skeletal abnormalities were seen in 4 of 6 patients.

Table. Clinical Presentation and L1CAM Mutation Details^a.

Proband ID	Sex	Ethnicity	Genetically screened at birth?	Proband GT	Father GT	Mother GT	Position (GRCh37)	Gene	OMIM in-heritance model	cDNA change	AA change	Meta SVM	MPC	Bravo MAF	Clinical presentation
KCHYD366-1	M	European	Yes	0/1	0/0	0/1	X:153133291:A:G	L1CAM	XLR	c.T1903C	p.W635R	D	1.99	0	AS, CH, bilateral adducted thumbs, white matter hypoplasia, epilepsy, macrocephaly, NDD, hypotonia, and skeletal and ophthalmological abnormalities
KCHYD268-1	M	African American	No	0/1	NA	NA	X:153134015:C:A	L1CAM	XLR	c.1546 + 1G>T	NA	NA	NA	0	AS, CH, NDD, and macrocephalya
KCHYD421-1	M	European	Yes	0/1	0/0	0/1	X:153135592:C:A	L1CAM	XLR	c.G910T	p.E304X	NA	NA	0	AS, CH, bilateral adducted thumbs, white matter hypoplasia, ACC, macrocephaly, NDD, esotropia, hypotonia, and skeletal and ophthalmological abnormalities
KCHYD498-1	M	European	No	0/1	0/0	0/1	X:153135842:C:G	L1CAM	XLR	c.806 + 1G>C	NA	NA	NA	0	AS, CH, bilateral adducted thumbs, white matter hypoplasia, ACC, SA, epilepsy, macrocephaly, NDD, cerebral palsy, and skeletal abnormalities
KCHYD553-1	M	European	No	0/1	NA	NA	X:153132173:C:A	L1CAM	XLR	c.G2362T	p.V788F	D	1.71	NA	CHb
KCHYD533-1	M	South Asian	No	0/1	0/0	0/1	X:153133452:C:T	L1CAM	XLR	c.1828 + 1G>A	NA	NA	NA	NA	AS, CH, ACC, cerebral palsy, epilepsy, NDD, spasticity, and a spinal fusion
KCHYD612-1	M	European	No	0/1	NA	NA	X:153134182:C:A	L1CAM	XLR	c.G1380T	p.W460C	D	1.58	NA	AS, CH, ACC, intracranial cyst, hypotonia, scoliosis, cerebral palsy, NDD, and skeletal and ophthalmological abnormalities

Abbreviations: AA, amino acid; ACC, agenesis of corpus collosum; AS, aqueductal stenosis; cDNA, complimentary DMA; CH, congenital hydrocephalus; GT, genotype; ID, identification; NA, not available; MAF, minor allele frequency; NDD, neurodevelopmental delay; SA, septal agenesis.

^aAvailable clinical information as well as demographic characteristics, mutation genotypes in proband and parents, mutation positions in GRCh37, functional annotations, prediction of deleteriousness by MetaSVM/MPC, and minor allele frequencies in the Bravo database (https://bravo.sph.umich.edu/).

^bNo other radiographic or clinical information available.

Figure. Transmitted and Unphased Mutations in L1CAM.

A, Available clinical neuroimaging phenotypes of congenital hydrocephalus probands with mutations in L1CAM. B, Mutation mapping of all identified L1CAM mutations. C, Pedigrees and sequencing electropherograms of Sanger sequencing depict L1CAM mutations in genomic DNA from congenital hydrocephalus probands. The mutation in proband KCHYD612-1 was not validated by Sanger sequencing owing to coronavirus disease 2019–related laboratory disruptions. Our Sanger validation rate is greater than 99%.

Discussion

Detection of these probands harboring L1CAM mutations suggests caregivers are missing opportunities to obtain a genetic diagnosis in patients with CH, with potential prognostic and therapeutic implications. This could result from a failure to recognize a syndromic phenotype, from a lack of awareness of available testing, and/or from the test used being incapable of identifying novel point mutations. Although our cohort size is limited and lacks ethnic diversity, these data highlight the diagnostic utility of WES in sporadic CH in contrast to other screening strategies. Given the discovery that rare mutations with large effect, including de novo mutations, account for a significant percentage of CH cases,¹ WES should be considered an essential diagnostic adjunct in the workup of all patients with sporadic perinatal CH.