Nephrotic syndrome genomic discovery in the Mass General Brigham Biobank identifies monoallelic MEFV variants as a risk factor for focal segmental glomerulosclerosis

Abstract

Introduction: Health system-based biobanks with genetic data provide a unique opportunity for nephrotic syndrome (NS) genomic discovery. This is predicated on finding cases in the electronic health record. Methods: We tested three strategies to identify focal segmental glomerulosclerosis (FSGS) or minimal change disease (MCD) cases in the 130,000 members of Mass General Brigham Biobank (MGBB). We analyzed a “synthetic proteinuria panel” of 192 Mendelian genes and the APOL1 kidney risk variants in those with exome sequencing (ES). We studied the associations between patients with Mendelian variants (MV), APOL1-HR genotype (APOL1) and outcomes. Validation of a novel gene-FSGS association was done in the Genomics England 100,000 Genome Project (100KG) and a global NS case-control cohort. Results: 319 MGBB participants had FSGS or MCD and ES data; reviewing pathology reports was the most accurate screening strategy. 31 (9.7%) of patients had MV and 24 (7.5%) had APOL1. 61% of genetic NS with a kidney biopsy report were classified as secondary FSGS. MV and APOL1 patients had a 3.1 (1.1–8.7) and 6.5 (1.3–32.3) increased odds of developing kidney failure, respectively. Unexpectedly, monoallelic pathogenic variants in MEFV (Mendelian gene for Familial Mediterranean Fever [FMF]) were found in 6 MGBB participants with FSGS, all of whom had features of collapsing glomerulopathy and thrombotic microangiopathy. 8 glomerular disease cases in the 100KG, unsolved via genome sequencing, had monoallelic pathogenic MEFV variants. Finally, a case-control study found a 3.8 increased odds of SRNS in individuals with pathogenic or likely pathogenic MEFV alleles (P = 7.8 × 10^–5). Conclusions: 17.2% of unselected adults with NS in the MGBB had a well-established genetic form, each associated with an increased risk of kidney failure. A biopsy read of secondary FSGS should not be used to rule out testing for genetic disease. Monoallelic pathogenic variants in MEFV may be a novel and underappreciated cause or susceptibility factor for SRNS/FSGS with distinct histologic features, even in the absence of clinical FMF.