Distinct genetic risk loci between biopsy-proven renal and non-renal lupus: a 10-year longitudinal cohort

Abstract Objective Systemic lupus erythematosus (SLE) is a heterogeneous disease which manifests as different subphenotypes. Distinct subphenotypes, such as lupus nephritis (LN), have been associated with increased genetic risk, but prior studies have been limited by cross-sectional and imprecisely subphenotyped cohorts. This study investigated the genetic basis for LN using a longitudinal cohort of distinctly subphenotyped patients. Methods SLE patients with biopsy-proven LN or never developed LN (NLN) were recruited from 8 tertiary referral centres. All patients had longitudinal clinical data for at least 10-years, or died during the study period. NLN patients had no renal involvement for at least 10-years. Subjects were genotyped and polygenic risk scores (PRS) were calculated using 230 SLE-associated SNPs. Genome-wide association analyses were also conducted for LN-vs-control, NLN-vs-control, and LN-vs-NLN comparisons, along with heterogeneity tests to assess differences in effect size. Results Among 1462 patients, 824 (56%) and 638 (43%) had LN and NLN, respectively. PRS was significantly higher in the LN cohort. Genome-wide significant variants were identified in HLA, TNFAIP3, BLK, and STAT4 loci specifically for LN patients, while STAT4 also remained significant for NLN patients. Direct LN-vs-NLN associations showed no statistically significant variants but heterogeneity tests revealed other genetic loci, including ELF1, OX40, DUSP22, and TPCN2. Conclusion Different subphenotypes of SLE are predisposed by distinct genetic risk loci, which can only be identified in clearly subphenotyped cohorts with sufficient longitudinal data. We identified unique genetic risk loci enriched among patients with biopsy-proven LN, and postulate that individual subphenotypes may have varying genetic predisposition.