Abstract Body: Background: Pediatric renovascular hypertension (PRVH), a rare disorder (1 in 1.9 million annually), is marked by renal artery narrowing and abdominal aortic coarctation, leading to renin-mediated hypertension. Its genetic basis is only partially explained by genetic syndromes, including Williams Syndrome (WS), which features supravalvular aortic stenosis (SVAS) and other vascular abnormalities. The ELN gene on chromosome 7q11.23 encodes elastin, an essential vascular connective tissue protein. ELN deletions are known to be linked to aortic stiffness and narrowing. Whether ELN gene variation is specifically associated with PRVH and whether copy number variation (CNV) or partial deletions of ELN underlie PRVH in individuals without a genetic syndrome is unknown.
Methods: Genotype intensities were analyzed using Illumina BeadArray v1.1, comprising 607,780 SNP probes with a median inter-SNP spacing of approximately 3 kb. The study compared 111 University of Michigan PRVH cases to 1,853 controls from the Michigan Genomic Initiative (1:19 ratio), matched by age, sex, and ancestry by principal components. CNVs were inferred using log R ratio (LRR) and B allele frequency (BAF) with a hidden Markov model. CNVRs were defined as CNV regions with at least 25 consecutive SNPs, merged within 50 Kb gaps, and analyzed for case-control associations using Fisher’s exact test.
Results: 627 unique CNVRs were identified, with the strongest PRVH-associated risk signal being a deletion (CN=1) at chr7q11.23 (72722981-74126034, GRCh37/hg19), in 5 PRVH cases and absent in all controls (P=5.29×10^-7). Genotype-phenotype analysis confirmed that all 5 individuals carried a clinical diagnosis of WS, and all individuals carrying a diagnosis of WS had 7q11.23 copy loss. ELN was deleted across all exons in all cases, with no evidence of smaller deletions. As genotype-phenotype correlations have been identified for non-cardiovascular manifestations of WS, we noted that in addition to ELN, the additional genes deleted in all five cases were BAZ1B, BCL7B, TBL2, MLXIPL, LIMK1, EIF4H, LAT2, RFC2, CLIP2, and GTF2IRD1. Vascular phenotypes showed abdominal aortic narrowing in 4 cases and isolated renal artery stenosis in one child, the latter a previously unappreciated WS vascular lesion.
Conclusions: Despite PRVH’s rarity, this study reports a statistically significant association between aortorenal dysplasia causing PRVH and a strongly deleterious genetic lesion of WS involving ELN deletions.