1012 Deep peripheral blood immunophenotyping identifies a subgroup of lupus nephritis patients characterized by high type 1 interferon signaling, persistent activated immune cells and poor renal response to standard of care at 1 year

Background

Lupus nephritis (LN) is a common and severe complication of systemic lupus erythematosus requiring renal biopsy to guide treatment decisions. Despite standard-of-care therapy, a third of patients with LN class III, IV, or V show a progressive decline in kidney function. Identifying distinct inflammatory processes associated with LN using non-invasive tools may improve treatment targeting.

Method

We applied mass cytometry using four 48-marker panels to characterize peripheral blood mononuclear cells from 140 patients with active, biopsy-proven proliferative (class III or IV +/- V, n=98) or membranous (class V, n=42) nephritis and 40 healthy controls in the Accelerated Medicine Partnership RA/SLE Network Phase II study. Renal response was determined at 1 year (n=107), and longitudinal samples were collected (n=49).

Results

By applying covarying neighborhood analysis, we observed a marked enrichment of cell neighborhoods that expressed type I interferon (IFN-I)-induced proteins, including MX1 (figure 1A) and Siglec1, reflecting a simple cytometric detection of an IFN-I signature. Specific immunophenotypic alterations, including activated and proliferating cells (Ki67hi) across immune cell types, such as plasmablasts (B8) and Tph (T11), were associated with IFN-I signaling (figure 1B). Amongst cells strongly associated with IFN-I, we identified a naive-like (CD45RA+CCR7+) CD4 T cell cluster (T12) expressing low levels of TCF1, suggesting early activation (figure 1C). We confirmed that IFN-I had an additive effect with TCR stimulation in decreasing TCF1 (TCF7 gene) expression in naive T cells using a recently published bulk RNAseq dataset1 (figure 2). Longitudinally, the cell subsets strongly associated with IFN-I, including T12, did not change in abundance over time independently from the renal outcome (longitudinal mixed effect model in complete responders p = 0.37, non-responder p = 0.62). As IFN-I signaling and the frequency of cell subsets varied across individuals, we stratified all samples using an unsupervised analysis based on 55 immune cell subsets. We identified 3 LN patient subgroups: the first group (G0) included all controls and 20% baseline LN patients; the two others (G1, G2) included only LN patients (46 patients each). G2 was associated with an increased proportion of GZMB+ T cells which were not correlated with IFN-I score (figure 1C). G1 displayed the highest cytometric IFN- I scores and was driven mainly by cells enriched with IFN-I-induced proteins, including T12 (figure 3A). Additionally, IFN-I cytometric score was strongly correlated with an IFN-I score in kidney leukocytes from paired samples (rho = 0.73, p < 0.001). Finally, after adjusting for a history of previous biopsy and immunosuppression at baseline, group G1 was significantly associated with non-response at 1 year (figure 3B).

Conclusion

IFN-I signaling in LN patients is associated with persistent enrichment of altered and activated immune cells, despite immunosuppression. Patients with increased IFN-I signaling in the peripheral blood have increased IFN-I signaling in the kidney and represent a group with poor response to standard-of-care.

Reference

Sumida T, et al. Type I interferon transcriptional network regulates expression of coinhibitory receptors in human T cells, Nature Immunology. 2022;23(4):632–42.