Melanie M Y Chan, Omid Sadeghi-Alavijeh, Seth du Preez, Catalin D Voinescu, Loes F M van der Zanden, Sander Groen In 't Woud, Michiel F Schreuder, Wout Feitz, Enrico Mingardo, Alina C Hilger, Heiko Reutter, Lisanne M Vendrig, Rik Westland, Glenda M Beaman, William G Newman, Adrian S Woolf, Horia C Stanescu, Adam P Levine, Detlef Böckenhauer, Daniel P Gale
The diagnostic yield for monogenic kidney disease was only 4.9% in this cohort, with common and low frequency variants potentially accounting for some of the missing heritability of CAKUT.
INTRODUCTION: Congenital anomalies of the kidneys and urinary tract (CAKUT) are the commonest cause of kidney failure in children and young adults. Over 50 monogenic causes have been identified. However, fewer than 20% of patients have a genetic diagnosis identified using targeted or whole exome sequencing. Here, we characterize the genomic architecture of CAKUT using whole genome sequencing (WGS).
METHODS: Based on WGS data from 1,052 unrelated individuals with CAKUT recruited to the UK's 100,000 Genomes Project, we used a panel-based framework to determine diagnostic yield and looked for gene-based enrichment of rare variants exome-wide. We performed sequencing-based genome-wide association studies (seqGWAS) and estimated the heritability attributable to common and low frequency variants.
RESULTS: The overall diagnostic yield was 4.9%, increasing to 7.4% in those with kidney agenesis/hypodysplasia and 11.1% in cystic kidney dysplasia. Family history (odds ratio (OR) 2.2; 95% confidence interval (CI) 1.1-4.4), consanguinity (OR 3.0; 95%CI 1.2-6.9) and extra-renal features (OR 3.1; 95% CI 1.7-5.7) independently predicted a monogenic diagnosis. Exome-wide rare variant and genome-wide common and low-frequency variant (minor allele frequency [MAF] 0.5% or more) association testing in a subset of 813 patients and 25,205 ancestry-matched control individuals identified a genome-wide significant association at 6q16.3 (rs117473527; OR 3.13; 95% CI 2.08-4.72; P = 4.8x10-8; MAF 0.01) requiring replication. Common and low-frequency variants were estimated to explain 23% (95% CI 1-45%) of the phenotypic variance observed in CAKUT although the confidence intervals were wide. A polygenic risk score for posterior urethral valves was constructed and validated in an independent European cohort of 77 cases and 2746 controls.
CONCLUSIONS: The diagnostic yield for monogenic kidney disease was only 4.9% in this cohort, with common and low frequency variants potentially accounting for some of the missing heritability of CAKUT.