OBJECTIVE:Type I interferons (IFN) play a key role in SLE pathogenesis, and an elevated IFN gene signature (IGS) has been associated with increased disease severity. This study aimed to retrospectively analyze the clinical and serological characteristics of SLE patients based on IFN-high or IFN-low status. METHODS:We analyzed a large cohort of 506 patients with SLE from the Toronto Lupus Clinic. Patients were classified as IFN-high or IFN-low based on IGS measured using the DxTerity Modular Immune Profile test. Demographic data, disease activity scores (SLE Disease Activity Index-2000 [SLEDAI-2K], Adjusted Mean SLEDAI-2K [AMS], Adjusted AMS Glucocorticoids [AMSG]), cumulative organ involvement, autoantibody profiles, and medication use were compared between high and IFN-low groups. RESULTS:Of the 506 patients, 291 (57.5%) were IFN-high and 215 (42.5%) were IFN-low. IFN-high patients were younger at study entry (median 46.3 vs. 54.2 years) and had shorter disease duration (median 14.1 vs. 22.7 years). IFN-high patients had higher disease activity scores (SLEDAI-2K, AMS, AMSG) and were more likely to be on glucocorticoids (38.5% vs. 27%) and immunosuppressants (63.6% vs. 45.6%), particularly mycophenolate (39.5% vs. 24.7%), and had a greater prevalence of positive autoantibodies. Despite higher disease activity, cumulative damage (SDI) was similar between IFN-high and IFN-low groups after multivariable analysis. CONCLUSIONS:Patients with an elevated IGS have more active and severe disease, accumulating more autoantibodies and requiring increased immunosuppression. Retrospective AMS/AMSG analyses further support IGS as a predictor of disease burden. Future studies should explore its role in guiding personalized treatment strategies.

Objectives SLE is characterised by unpredictable flares and periods of quiescence. Currently, the mechanisms driving flares remain incompletely understood. Type I interferon (IFN) signalling is strongly implicated in SLE pathogenesis and promotes expansion of age-associated B cells in murine lupus, which are phenotypically similar to double negative 2 (DN2) B cells in humans. We investigated how IFN exposure influences B cell activation and differentiation in SLE, and whether type I IFN promotes DN2 differentiation in human naïve B cells in vitro.Methods A cytometry by time of flight panel was developed to identify B cell subsets, markers associated with B cell activation and IFN-stimulated proteins (ISPs), with a composite ISP score used as a surrogate for IFN exposure. Peripheral blood mononuclear cells from 18 healthy controls (HCs), 25 quiescent patients and 41 recently flaring patients with SLE were analysed. HC naïve B cells were stimulated under conditions that promote or inhibit DN2 differentiation with and without IFNs.Results All B cell subsets exhibited higher ISP levels in flaring than quiescent disease. DN2 cells were more abundant in flaring patients and correlated with the ISP score. ISP scores strongly correlated with markers associated with B cell activation even after controlling for disease status. In vitro, IFNα and IFNβ promoted DN2 differentiation in HC B cells. In patients treated with the type I IFN receptor-blocking therapy anifrolumab, DN2 frequency and ISP scores significantly decreased.Conclusions Type I IFN promotes human B cell activation and DN2 differentiation, supporting IFN blockade as a strategy to reduce pathogenic B cells and prevent SLE flares.

Objectives To characterize longitudinal changes in antinuclear antibody (ANA) titers and patterns, together with the autoantibody profile, in asymptomatic ANA-positive individuals and undifferentiated connective tissue disease (UCTD) patients, and to identify serological predictors of systemic autoimmune rheumatic diseases (SARD) progression. Methods We analyzed autoimmune serology from 225 asymptomatic ANA-positive or UCTD subjects with longitudinal follow-up (1-7 years) to assess changes over time, seroconversion, and progression to (SARD). ANAs were quantified by indirect IF using the Kallestad® HEp-2 kit and specific autoantibodies measured using the Bioplex® 2200 ANA Screening System which assesses the levels of anti-dsDNA, -chromatin, -Ro, -La, -Sm, -SmRNP, -RNP, -Jo-1, -Scl-70, -centromere and -ribosomal P antibodies. Results At baseline, a high ANA titer (≥1:640) predominated in 60% of subjects, with speckled and homogeneous patterns representing the majority of immunofluorescent patterns. The most prevalent baseline autoantibodies were Ro (23.3%) and RNP (15.7%). Among subjects with serial ANA testing (n=77), 60% had initial high titers (≥1:640). Upon follow-up, 53% had declining titers with 6.5% seroconverting to ANA-negative. Of those who seroconverted, 60% had high titers (≥1:640) initially and 40% had a dense fine speckled (DFS) pattern. The median time to ANA loss was 3.3 years. Among the 6 subjects who ever demonstrated a DFS pattern, none of whom progressed to SARD, 33.3% converted to negative, and 33.3% transitioned to other ANA patterns. This variability underscores the dynamic but low-risk nature of DFS pattern. Among the 135 subjects with serial autoantibody profiling, Ro remained the most prevalent antibody (40%) and was usually stably elevated (89% remained persistently positive), showing minimal fluctuation over time. In contrast, RNP (17%) and dsDNA (6.7%) showed the greatest instability, with 44% of dsDNA-positive and 22% of RNP-positive subjects becoming negative. Interestingly, Scl-70, which was positive in 9 patients, also showed instability, with 4 of these patients losing positivity over time. Nine patients progressed to a defined SARD after a median of 31 months. Overall, progressors had high ANA titers (≥1:320), and none became ANA-negative. Among this group, dsDNA (89%) and Ro (SSA) (56%) were the most prevalent autoantibodies. Conclusion Many subjects demonstrated immunological improvements. Subjects who achieved ANA negativity typically began with high ANA titers that declined over time. In contrast, progressors showed persistently high ANA titers, frequently accompanied by rising dsDNA and stable Ro reactivity. These findings support the clinical importance of serial ANA and ENA monitoring to differentiate transient autoimmunity from evolving systemic disease.

Objectives Up to 65% of patients with systemic lupus erythematosus (SLE) develop lupus nephritis (LN), a major cause of renal failure. Approximately 30% of LN patients fail to respond to standard immunosuppressive therapy, emphasizing the need for biomarkers that predict therapeutic response at flare onset. Interferon-induced gene (IFI-G) expression is a promising candidate biomarker, as elevated IFI-G levels in blood and renal tissue have been linked to more severe disease. We have previously shown that IFI-Ps (ISG15, MX1, IFNAR1) serve as reliable surrogates for IFI-G expression. This study aimed to develop and validate an imaging mass cytometry (IMC) panel to spatially characterize IFI-Ps and immune cell subsets in LN kidney biopsies, and to explore whether IFI-P expression correlates with treatment response. Methods We optimized an IMC panel to simultaneously evaluate IFI-P expression and characterize immune cell subsets directly within kidney tissue. Paraffin-embedded biopsies from LN patients in the Lupus Nephritis New Emerging Team and University of Toronto Lupus Clinic cohorts were analyzed. A pseudo-tissue composed of IFN-stimulated and unstimulated peripheral blood mononuclear cells in a clot that was embedded paraffin was used as a positive and negative control for antibody specificity. The panel includes 34 metal-conjugated antibodies to detect renal resident cells, infiltrating immune cells (T cells, B cells, monocytes, macrophages, dendritic cells), IFI-Ps, and fibrosis. Results The IMC panel successfully detected renal and immune cell subsets, as well as distinct patterns of IFI-P expression in various renal compartments (Figure 1A). IFI-P expression demonstrated a moderate to strong correlation between the different kidney compartments (Figure 1B). Preliminary analysis of 10 LN biopsies revealed a trend to increased IFI-P staining intensity in all of the renal compartments of nonresponders when compared to responders (Figure 1C). This achieved statistical significance (p < 0.05) for IFNAR1, MX1, and PKR in the glomerulus and PKR in the nonproximal tubules, consistent with the concept that elevated renal levels of IFN are associated with poorer response to treatment. Figure 1. Spatial profiling of interferon-induced proteins in LN kidney biopsies by imaging mass cytometry (IMC). (A) Representative IMC images of kidney biopsies from a treatment responder and non-responder, showing expression of ISG15 (yellow). MX1 (cyan). PKR (green), and IFNAR1 (red) overlaid with DNA (blue). Non-responders exhibited stronger and more diffuse IFI-P staining across renal compartments. Scale bar = 100 μm. (B) Correlation of MX1 expression between kidney compartments demonstrates coordinated interferon activity within tissue. MX1 is shown as a representative IFI-P, with similar correlations observed for the other IFI-P markers. Strong correlation was observed between proximal and non-proximal tubules (R = 0.84. p = 0.0045) and a moderate correlation between proximal tubules and glomeruli (R = 0.64. p = 0.054). (C) Quantitative comparison of mean IFI-P intensity across kidney regions in responders (n = 6) and non-responders (n = 4). Non-responders showed higher expression of IFNAR1 (p= 0.00952), MX1 (p= 0.0381), and PKR (0.0190) in the glomerulus, and increased PKR expression in the non-proximal tubules (p= 0.0381). Asterisks (*) denote statistically significant differences between treatment groups. Conclusion This study establishes a validated IMC-based approach for high-resolution spatial profiling of interferon signatures and immune cell subsets in LN kidney biopsies. Preliminary data suggest that elevated IFI-P expression is associated with nonresponse to therapy, supporting further investigation of IFI-Ps as predictive biomarkers. Application of this panel to larger LN cohorts may enable early patient stratification and guide precision treatment strategies to improve renal outcomes.

Objectives Systemic lupus erythematosus (SLE) is characterized by unpredictable flares interspersed with periods of disease quiescence. Elevated interferon (IFN) levels increase the likelihood of flares, but the precise immunologic mechanisms by which this occurs are unclear. In mice, IFN exposure expands age-associated B cells (ABCs), a population enriched for autoreactive and ANA-secreting cells. In this study, we examined the role of IFN in the activation and differentiation of B-cell subsets in flaring and quiescent SLE patients. Methods A CyTOF panel was developed to quantify IFN-induced proteins (IIPs) across peripheral blood immune populations. A composite IIP score (mean expression of 6 IIPs) was generated for each cell population as a surrogate for IFN exposure. 15 healthy controls (HCs), 26 quiescent (clinical SLEDAI > 0 for 1 year with no increase in immunosuppressive treatment, ≤ 10 mg prednisone) and 42 recently flaring (<1 month, change in clinical SLEDAI-2K ≥ 1 requiring escalation of therapy) were analyzed. To assess the direct effects of IFN, naïve B cells from HCs were isolated and stimulated with IFNα, IFNβ, or IFNγ. Cells were cultured under conditions that either promoted or inhibited ABC differentiation, including IL-21, anti-CD40, Fab2, CpG, or IL-4. Results CyTOF identified 7 B cell subsets, all of which exhibited higher IIP levels and greater activation in flaring vs quiescent patients (Figure 1A). ABCs were more abundant in flaring patients, and their frequency correlated with the global IIP signature (Figure 1B,C). Expression of activation markers (CD86, TLR7, TLR9, HLA-DR, Ki67) was strongly associated with IIP score, but not disease status, indicating that IFN, rather than flare alone, drives B cell activation (Figure 1D). Importantly, the association between activation and IFN exposure was evident even within individual patients, where the top 10% of IFN-experienced B cells had significantly higher activation than the bottom 10%. In vitro, IFNα and IFNβ directly induced expression of activation markers within 18-24 hours. In isolated naïve B cells, all 3 IFNs increased ABC differentiation, even without canonical ABC-inducing signals (Figure 1E). Notably, IFN overcame IL-4-mediated suppression of ABC differentiation, in part by reducing IL-4Rα and inducing TLR7 expression. In SLE patients treated with the IFN-blocking therapy Anifrolumab, ABC frequency and IIP signatures decreased (Figure 1F). Figure 1. A) IIP score for HCs, quiescent, and flaring SLE patients in B cell subsets. Higher IIP scores were found in flaring versus quiescent, and SLE patients versus HCs. (Mann Whitney U test with BH correction for multiple tests) B) Frequency of B cell subsets displayed as the proportion of CD19+ B cells. Flaring patients had more ABCs than quiescent patients and HCs. (Mann Whitney U test with BH correction for multiple tests) C) Correlation between cellular abundance and IFN signature. The proportion of ABCs were correlated with IIP score as well as IFN-stimulated genes (ISGs). IIP score also correlated with the proportion of PBs. (Spearman correlation) D) Correlation of IIP score and markers of activation. The expression of activation markers correlated with IIP score for SLE patients. (Spearman correlation) E) ABC differentiation from purified naïve B cells. 5 days of incubation with IFNα, IFNβ, or IFNγ caused significantly more differentiation of ABCs, irrespective of the incubation conditions (+/− IL-21, IL-4, anti-CD40, CpG). Results are displayed as the fold change from the respective non-IFN conditions. (Student’s T test with Holm correction for multiple comparisons post Shapiro-Wilk test to assess for normality) F) The ABC profile of patients treated with Anifrolumab versus standard-of-care. Both the IIP score and proportion of ABCs were significantly reduced in Anifrolumab-treated patients. (Mann Whitney U test) In Anifrolumab-treated patients, the frequency of ABCs showed a trend to correlation with the IIP score. (Spearman correlation) Conclusion IFN exposure drives human B-cell activation and promotes differentiation of ABCs. Our results identify a mechanistic link between IFN signaling and pathogenic B-cell development, and support IFN blockade as a strategy to reduce pathogenic ABCs and prevent SLE flares.

Objectives Systemic Lupus Erythematosus (SLE) is a chronic autoimmune disease associated with a severe morbidity and mortality. Around 70% of SLE patients follow a relapse-remitting pattern of disease characterized by flares of disease activity, followed by prolonged periods of disease quiescence. Memory CD4+ T cell subsets have been shown to play an important role in driving the autoantibody production which causes flares in SLE, however the precise T cell changes that accompany flares are unknown. Methods CITE-seq and TCR-seq were performed to assess the transcriptomic profiles of CD4+ memory T cells in flaring and quiescent SLE patients. CD4+ memory T cells were isolated from PBMCs by negative selection using magnetic sorting, stained with oligo-conjugated antibodies against surface proteins for subset classification, and subsequently partitioned, barcoded, and sequenced. We examined samples from 15 distinct patients at 2 separate clinical visits spaced one year apart, yielding 30 samples. The longitudinal nature of our data allows us to inspect transcriptional changes both between and within patients. Results Integrated analysis of 30 samples identified 10 immune cell clusters (Figure 1A). At baseline, flaring patients (n=9) were significantly enriched for Tfh, Th2, Th17 cells, and a Treg subset, while quiescent patients (n=6) had increased Th1 cells. TCR repertoire analyses at baseline revealed a higher proportion of expanded clonotypes in flaring patients, which was not seen in quiescent patients. Interestingly, we also found that there was a higher proportion of expanded clonotypes at follow-up in various subsets of interest, particularly in flaring patients that later became quiescent, suggesting tissue egress and recirculation following resolution of inflammation. Clonal overlap among subsets was markedly greater in flaring patients, suggesting shared antigen specificity and differentiation from common progenitors. More specifically, we identified 2 functionally deviated/exhausted Treg subsets (ISGhi/ISGlo) (Figure 1B) and, at baseline, found notable clonal overlap between the ISGhi Treg subset and Th2/17 cells and between the ISGlo subset and Tfh/Tph cells in flaring patients, which was absent in quiescent patients (Figure 1C). This suggests that there are 2 distinct subsets of cells with shared antigen exposure and/or functional plasticity; one that is exposed to an IFN-rich environment in the tissue, and another that is more involved in T-B cell interactions within lymphoid compartments. Conclusion We found abnormal Treg subsets with features of exhaustion and functional deviation that shared antigen specificity with other T helper cells. Their increased prevalence during flares suggests that dysregulated immunoregulation may contribute to SLE pathogenesis.

Objective This study examined serum brain injury markers and their associations with disease features, cytokines associated with microglial activation in lupus and cognitive dysfunction (CD) in adolescents with childhood-onset SLE (cSLE).Methods We used cross-sectional data from cSLE patients (aged 12–17 years) and age-matched, sex-matched healthy controls. Serum levels of brain injury markers (serum neurofilament light, glial fibrillar acidic protein (GFAP), Tau), interferon (IFN)-α, IFN-γ and interleukin-6 (IL-6) were quantified using Simoa assays. cSLE features included disease activity (Systemic Lupus Erythematosus Disease Activity Index 2000), damage (Systemic Lupus International Collaborating Clinics damage index) and glucocorticoid (GC) exposure. A neurocognitive battery assessed executive function, attention and working memory, and CD was determined using standardised scores. We compared brain injury marker levels between cSLE and controls, and those with and without CD using Wilcoxon rank-sum tests. We calculated correlations between injury markers, disease features and cytokines and examined differences in disease features between those with and without high-level brain injury markers (>90th percentile) (using Bonferroni correction).Results Participants included 56 cSLE patients (median disease duration=10.6 months (IQR 2.0–14.1), one with neuropsychiatric lupus) and 43 controls. Levels were higher in cSLE versus controls for GFAP (z=−3.97, p<0.001), Tau (z=−2.10, p=0.035), IFN-α (z=−4.80, p<0.001), IFN-γ (z=−2.42, p=0.015) and IL-6 (−3.09, p=0.002). Severe CD (≥2 SD from standardised mean) was present in 31% cSLE versus 9% controls (chi2=6.69, p=0.01), associated with higher Tau levels for cSLE (z=−3.94, p<0.001). High-level brain injury markers were observed in 13 (23%) cSLE patients associated with higher SLEDAI-2K, IL-6 levels and current GC dose.Conclusion Brain injury marker levels were high and associated with disease activity and CD in this cSLE adolescent cohort, suggesting a link between systemic inflammation and clinically under-detected neuronal/glial injury. Larger, longitudinal studies should explore the potential clinical utility of brain injury markers for clinical assessment of brain involvement in cSLE.

Objectives Systemic autoimmune rheumatic diseases (SARD) exhibit a prolonged pre-clinical phase during which they have anti-nuclear antibodies (ANAs).[1] However, ANAs cannot be reliably used to predict impending disease because a subset of healthy women are ANA+ (~20%) and the majority of these individuals will not progress to SARD.[2] Why some individuals progress while others remain asymptomatic is unknown. Our objective is to evaluate functional alterations in innate immune populations during SARD development. Methods CITE-Seq was conducted on innate immune cells from 24 patients including healthy controls (HC, n=5), non-progressor (NP, n=6) IFN high or low, progressors (P, n=5) or SARD (n=8) patients. Differential gene expression analysis was performed to identify genes of interest. Spectral flow cytometry and plasma ELISAs were conducted in an expanded group of patients to validate differences in CITE-Seq genes of interest. A monocyte cell line (THP-1) was used to investigate the kinetics of 1 gene identified by CITE-Seq to gain a better understanding of potential functional implications. Results Non-progressors (NP) exhibited increased gene expression of heat shock proteins (HSPs) like HSP70 and CD52 compared to P (Figure 1A). Both proteins are proposed to promote immune regulation through tolerogenic effects on innate immune cells. Conversely, P exhibited increased gene expression of MHC class II alleles, which are associated with immune activation. Using flow cytometry, we confirmed the differences between groups of surface expression of CD52 and MHC class II on innate immune cells (Figure 1B). Little is known about how HSPs are regulated and expressed. To better understand the kinetics of HSP70, THP-1 cells were heat shocked. Gene expression showed rapid and transient upregulation which was attenuated by 18h, while soluble HSP70 increased steadily following activation of the heat shock response (Figure 1C). Protein HSP70 was decreased at the 12h timepoint using immunofluorescence (Figure 1D), following heat shock, suggesting the release of HSP70. We therefore measured plasma HSP70 in our patient cohort and found that soluble HSP70 was increased in NP compared to P (Figure 1E), in support of our CITE-Seq results. Ongoing experiments are being conducted to evaluate the release of HSP70 by purified innate immune cells from our patient cohort. Figure 1. Innate immune cells during the progression of SARD. A) Feature plots from CITE-Seq of monocytes and DCs shotting HSPA1A (HSP70), CD52 (CD52) and HLA-DRB5 (MHC Class II). B) Spectral flow cytometry showing representative results from classical monocytes tor HLA-DR and CD52 protein expression. C) Gene expression of HSPA1A (HSP70) by RT-qPCR and soluble HSP70 (sHSP70) concentration by supernatant ELISAs measured in a THP-1 monocyte cell line. Cells were heat shocked (43°C for 1h) and subsequently rested for 6h, 12h, 18h, 24h before collection. E) Heat shocked THP-1 cells were rested for 12h and used for cytospin with subsequent immunofluorescence staining with DAPI for cell nuclei and HSP70 (scale bar = 50 μm ). E) Plasma ELISAs for soluble HSP70 in patient cohort. Data was analyzed using a Kruskal-Wallis test with Dunn post hoe and Bonferroni correction (* = p-value < 0.05). Conclusion Our data shows that NPs exhibit mechanisms of immune suppression that are decreased in P. Importantly, P exhibits immune dysregulation prior to clinical progression. These results will allow us to further investigate the immunological differences in innate cells that may drive or inhibit progression in SARD. References [1.] Goldblatt F. Lancet 2013;382:797-808. [2.] Wither J. Arthritis Res Ther 2017;19:41.

OBJECTIVE:To determine if the levels of five urinary biomarkers (UBs), including cluster of differentiation 163 (CD163), monocyte chemoattractant protein-1 (MCP-1), adiponectin, soluble vascular cell adhesion molecule and platelet factor 4 (PF4), measured 24 months after a lupus nephritis (LN) flare are associated with adverse long-term outcomes. METHODS:We included patients with an LN flare who had a preflare estimated glomerular filtration rate (eGFR) ≥60 mL/min and stored urine 24±3 months after the flare. The following outcomes were then examined: (1) time to a subsequent LN flare and (2) time to 30% sustained decline in eGFR. UBs were measured by ELISA 24±3 months after the LN flare. The results were normalised to urine creatinine and expressed as pg per mmol of urine creatinine. RESULTS:69 patients with LN were included, the median (IQR) follow-up time after their 24-month urinary sample collection was 129 (97.5-150) months. 50 patients achieved a primary efficacy renal response 24 months after the LN flare. This subcohort of patients had significantly lower UB levels. In this subcohort, 27 (54%) experienced a subsequent LN flare with a median time to flare (IQR) of 3.5 (1.67-6.87) years, and 10 (20%) had a 30% decline in eGFR at a median time of 4.38 (3.73-5.33) years after their 24-month urinary sample collection. Elevated levels of MCP-1 (HR 1.40 (1.11-1.76), p=0.004) and CD163 (HR 1.14 (1.00-1.38), p=0.01) predicted a subsequent LN flare. While CD163 (HR 1.16 (1.02-1.32), p=0.02), MCP-1 (HR 1.33 (1.01-1.74), p=0.04), adiponectin (HR 2.67 (1.68-2.46), p<0.001) and PF4 (HR 1.14 (1.04-1.25), p=0.002) predicted a 30% decline in eGFR. CONCLUSION:UBs measured 24±3 months after an LN flare were associated with subsequent flares and a clinically meaningful decline in kidney function.

OBJECTIVES:Arthritis is a common manifestation of systemic lupus erythematosus (SLE). We examined the prevalence and associations of arthritis subtypes in SLE, focusing on associations with type I interferon (IFN) scores. METHODS:In this observational cohort study at the University of Toronto Lupus Clinic (July 1970-August 2024), arthritis was defined clinically as non-deforming arthritis (NDA), Jaccoud's arthropathy (JA) and arthritis with clinically irreducible deformities (CIDA). Univariable and multivariable (MV) logistic regression was used to assess associations with arthritis subtypes with NDA as reference in the prevalent cohort and subgroup with available IFN scores (reported as high/low). RESULTS:Among 2264 patients, 1248 (55.1%) had arthritis: 908 (72.8%) had NDA, 239 (19.2%) JA, and 101 (8.1%) CIDA. In the multivariable logistic regression analysis, JA was associated with longer disease duration (odds ratio [OR] 1.05 [95% CI: 1.03, 1.08]) and female sex (2.06 [1.11, 3.83]) whereas CIDA was associated with neurological involvement (1.79 [1.13, 2.84]), anti-Ro antibodies (1.71 [1.01, 2.90]), higher adjusted mean joint count (1.09 [1.03, 1.77]) and lower adjusted mean SLEDAI-2K over follow-up (0.89 [0.81, 0.98]) compared with NDA. In patients with available IFN scores (n = 475), CIDA was associated with a low IFN signature in unadjusted analysis, whereas both JA and NDA had a high IFN signature. CONCLUSION:Distinct clinical and serological patterns differentiate JA and CIDA, suggesting unique underlying mechanisms of deforming arthritis in SLE. Both NDA and JA exhibited a high IFN signature, while CIDA showed a higher joint burden, lower disease activity and lower IFN signature relative to NDA.

PV036 / #672 Poster Topic: AS04 - Biomarkers Type I interferons (IFN) are pivotal in the pathogenesis of SLE, with studies showing high IFN gene signature (IGS) status associated with certain organ manifestations, autoantibody profiles, and disease severity. With novel medications targeting the interferon pathway, an enhanced understanding the IGS in patients with SLE is necessary. In this study, we investigated the differences between IGS levels in relation to clinical characteristics in a large cohort of patients with SLE. Patients meeting 2019 EULAR/ACR classification criteria for SLE from a single center were included. Whole blood collected cross-sectionally was analyzed for IGS by the DxTerity assay, categorizing patients into IFN high or IFN low status. The SLICC/ACR damage index (SDI), antibody status, glucocorticoid use, and use of immunosuppressives were analyzed according to IFN status. Additionally, the SLEDAI-2K, SLEDAI-2KG, and clinical SLEDAI-2K were characterized cumulatively from 5 years prior to whole blood collection to last available visit, with adjusted mean SLEDAI (AMS) and AMS-G calculated from the previous 10 years and stratified based on IFN status. In total, 506 patients with a median age of 49.5 years (IQR 37.26-60.49) were included, with 91.5% female. Overall, 291 (57.5%) were IFN high and 215 (42.5%) were IFN low. The median disease duration was longer in the IFN low group (22.7 years, IQR 11.58-21.05) than in the IFN high group (14.06 years, IQR 7.90-24.71) (p<0.001). The median SLEDAI-2K score was higher in the IFN high group (2.0, IQR 0.00-4.00) then in the IFN low group (0.0, IQR 0.00-4.00) (<0.001), as was the SLEDAI-2KG (8.0, IQR 0.00-3.75 vs 6.0, IQR 0.00-2.00) (p<0.001) though the clinical SLEDAI-2K was not significantly different. Similar trends were seen with the cumulative AMS (3.94, SD=2.61 vs 3.27, SZD 2.47) (p=0.004) and AMS-G (5.36, SD=3.53 vs 4.41, SD=3.25) (p=0.002). There was no difference in the proportion of the SLEDAI-2K organ domains between the IFN high and low groups, other than the hematologic domain (IFN high 82.8% vs. IFN low 73.0% [p=0.011]). The SDI was similar between the 2 groups. More patients in the IFN high vs. low group had positive autoantibodies (79.7% vs 67% [p=0.002]), including Smith (56.7% vs 32.6% [p<0.001]), RNP (66.7% vs 49.3% [p<0.001]), Ro (67.4% vs 47.0% [p<0.001]), La (30.9% vs 17.2% [p=0.001]), chromatin (75.6% vs 41.9% [p<0.001]), dsDNA (61.2% vs 38.1% [p<0.001]) and ribosomal P (27.1% vs 7.9% [p<0.001]) autoantibodies. There were no differences in levels of C3 or lupus anticoagulant, but more patients in the IFN high group had low C4 and positive anticardiolipin. More patients with high IFN were on glucocorticoids (38.5%) then were patients with low IFN low status (27%) (p=0.009). More IFN high status patients were on immunosuppressive agents (185; 63.6%) vs the IFN low group (98; 45.6%) (p<0.001). Similar trends were seen in our inception cohort, defined as those patients evaluated within 1 year of diagnosis. In this large cohort of patients with SLE, IGS status may help to predict overall disease severity as per SLEDAI-2K, SLEDAI-2KG, AMS, and AMS-G (though not the clinical SLEDAI-2K), use of glucocorticoids, and overall use of immunosuppressive therapy, and is associated with more autoantibody positivity. IFN level did not reliably predict presence of specific SLEDAI-2K organ domains or damage. More studies are needed to assess those clinical characteristics associated with an elevated IGS and to determined who may respond best to type I IFN targeted therapies.

O021 / #624 Topic:AS22 - SLE Heterogeneity ABSTRACT CONCURRENT SESSION 03: INNATE AND ADAPTIVE IMMUNITY IN SLE 22-05-2025 1:40 PM - 2:40 PM High levels of peripheral blood interferon (IFN)-induced gene (IIG) expression are a characteristic feature of SLE and associated with an increased risk of flare. However, how these global changes correlate with those in individual immune populations and act to promote flares remains unclear. To address this question, we examined the IFN-induced immune changes in SLE patients at a single-cell level. A 40-marker CyTOF panel was used to measure IFN-induced protein (IIP) levels in the peripheral blood immune populations of 15 healthy controls (HC), 26 quiescent (clinical SLEDAI-2K = 0 for 1 year), and 42 recently flaring (clinical SLEDAI-2K ≥ 1 requiring an escalation of therapy) SLE patients. Twenty-nine immune populations were identified (Figure 1A). The mean IIP levels in all populations strongly correlated with global IIG expression, and were higher in flaring than quiescent patients (Figure 1B,C). Despite this correlation, there was significant heterogeneity in IIP expression between and within the cell subsets of individual patients, with the highest median levels of IIP seen in monocytes, plasmablast/plasma cells, and activated double positive T cells. These differences paralleled the response of these populations to exogenous IFN in HC cells in vitro. Within each cell subset of individual patients, there was a variably broad distribution of IIP expression, sometimes with distinct peaks (Figure 1D). To assess the factors contributing to this heterogeneity, we performed an analysis of extremes comparing the top and bottom 10% of IIP expressing cells in each subject (Figure 2A). Although the top IIP expressing cell subset of most populations had elevated levels of activation markers, such as Ki67, CD86, TLR7, TLR9, and HLA-DR, these molecules were induced by IFN in vitro, suggesting that IFN plays a direct role in their upregulation in vivo (Figure 2B). Notably, increased levels of the trafficking markers were also seen in the high IIP expressing cell subset, but with the exception of β7 (an integrin implicated in homing and retention in the gut), were not induced by IFN in vitro. Furthermore, these trafficking molecules demonstrated distinct patterns of expression, suggesting that these cells had transited different tissues. Longitudinal analysis of IIP expression over time revealed relatively stable levels despite changes in disease activity, and although the levels of IIP in the different cell subsets tended to correlate with each other, only the levels within B cells were associated with sustained or recurrent disease activity 1 year later. Figure 1. A) UMAP of the individual cell types showing their differential abundance and relatedness, as well as the average of 6 IIPs In HCs, quiescent and flaring patients: 29 cell types were identified based on their expression of the markers in our panel. There is a gradient in expression of average IIP expression in most cell types from low to high in HCs, quiescent patients, and flaring patients. B) Correlation matrix in all cells: Correlation between IIP expression (shown on theyaxis) and IIG expression in individual cell populations (shown on thexaxis). R values are denoted by colour, and p values by the size of the dots. C) Immunologic differences in IIP expression in immuno cell populations comparing flaring and quiescent patients: Waterfall plot showing the differential levels of IIP scores between flaring and quiescent, with bars above the line indicating increased expression in SLE patients. 27/29 immune cell subsets have significantly higher IIP scores in flaring patients relative to quiescent. D) Heterogeneity in IIP expression levels within the cell subsets of individual patients and HCs: Patients were separated into IIG high and low groups based on the top and bottom 15% of IIG score. Regardless of IIG group, there was heterogeneity in the IIP signature on a single cell level that was found within patient cells. Myeloid cells had the most marked heterogeneity, followed by T cells and then B colls. Figure 2. A) Analysis of extremes. Comparing the top and bottom 10% of IIP expressing cells within the same HCs and patients from ex vivo samples, it was found that certain activation and trafficking markers are upregulated in the IIP high cells, some of which are directly induced by IFN. R values are denoted by colour, and p values by the size of the dots. B) Incubation with IFNα and IFNβ induces several of the cellular markers that are associated with increased IIP expression in-vitro. PBMCs from healthy controls were stimulated with the indicated IFNs for 18 or 24 hours in the presence of Golgi-Stop for the last 2 hours. Shown are fold increases relative to unstimulated control. Although the mean IIP expression in each immune population correlates strongly with the IFN signature, there is significant heterogeneity between and within the cell populations of each patient in IIP expression. This appears to result not only from variability in the cells capacity to respond to IFN, but also variable exposure to IFN as cells traffic through the body.

O045 / #659 Topic: AS05 - CNS Lupus ABSTRACT CONCURRENT SESSION 07: COGNITION IMPAIRMENT IN SLE – RECENT ADVANCEMENT AND EMERGING RESEARCH 23-05-2025 1:40 PM - 2:40 PM Cognitive impairment (CI) is a common manifestation in patients with systemic lupus erythematosus (SLE). Despite its impact on patient quality of life, treatments remain limited as its pathogenesis is poorly understood. The Automated Neuropsychological Assessment Metrics (ANAM) has superior patient acceptability and feasibility in ambulatory settings compared to the American College of Rheumatology Neuropsychological Battery (ACR-NB) [gold-standard test] and is validated in screening for CI in SLE. Data from our laboratory have revealed that serum S100A8/A9 and MMP-9 are associated with CI measured by the ACR-NB. However, the relationship between these serum analytes, ANAM subtests and CI has not been elucidated. We therefore aimed to determine if serum analytes are associated with CI measured by the ANAM. We cross-sectionally analyzed the data of 327 adults aged 18-65 who were followed longitudinally between January 2016 and October 2019 at a single SLE center. All participants fulfilled the 2019 EULAR/ACR SLE classification criteria. Cognitive function was measured using ANAM throughput scores, and serum levels of 9 analytes (IL-10, IL-6, IFN-γ, TNF-α, TWEAK, S100B, S100A8/A9, NGAL and MMP-9) were measured using ELISA. The K-means clustering algorithm was used to cluster the patient data, and the Principal Component Analysis (PCA) characterized the clusters. The silhouette coefficient(s) was calculated for 2 to 15 clusters to determine the optimal number of clusters. PCA identified 2 principal components explaining 36.2% of the variance in ANAM throughputs and serum analytes. The first component (26.7% of the variance) was correlated with ANAM throughputs, with the strongest contribution from procedural reaction time. The second component (9.44% of the variance) was correlated with serum analyte measurements, with the strongest contribution from TNF-alpha. (Figure 1) The highest silhouette value was found for 2 (s = 0.177) and 3 (s = 0.176) clusters. Only 4% of patients were classified in the 3 cluster model, so a 2 cluster model was selected. Cluster 1 had low throughput scores representing CI, and Cluster 2 had higher throughput scores representing no CI. A significant difference was observed in mean serum S100A8/A9 (SMD = 0.362), MMP-9 (SMD = 0.178) and IL-6 (SMD = 0.311) between the clusters, reflected by their correlation with the first principal component. (Figure 2) Serum levels of S100A8/A9, MMP-9 and IL-6 had a strongly negative correlation between the Go No Go and Running Memory throughputs. Figure 1: Correlation matrix for individual ANAM throughputs and serum analyte levels Figure 2: Biplot of the first 2 principal components, with 2 clusters and association with analytes Serum S100A8/A9, MMP-9, and IL-6 are associated with CI in SLE as measured by the ANAM. Patient clusters with elevated serum S100A8/A9, MMP-9, and IL-6 had strongly negative associations with throughputs representing impairment in executive function, simple attention and processing speed. Further studies are needed to uncover mechanistic relationships between these analytes and CI in SLE, and whether they may represent valuable therapeutic targets for further exploration.

Renal involvement in systemic lupus erythematosus (SLE), lupus nephritis (LN), is common and can result in significant morbidity, including progressive renal dysfunction, and even ultimately leading to death. LN is heterogeneous complicated by the immunologic component, and it is critical to accurately classify LN to direct optimal therapy. Accordingly, identification of objective markers is paramount in reflecting disease stage and monitoring treatment response. In part two of this series, we comprehensively examine LN disease classification, therapies and potential markers to guide therapeutic options.

O020 / #593 Topic:AS16 - Lupus Nephritis-Pathogenesis ABSTRACT CONCURRENT SESSION 03: INNATE AND ADAPTIVE IMMUNITY IN SLE 22-05-2025 1:40 PM - 2:40 PM 20-65% of patients with systemic lupus erythematosus (SLE) will develop lupus nephritis (LN), with up to 30% failing to respond to standard immunosuppressive therapy. These patients are at risk of kidney functional decline, highlighting the need for biomarkers that predict therapeutic response at flare onset. One potential biomarker is interferon-induced gene (IFI-G) expression. Higher levels of IFI-G expression in the peripheral blood have been associated with a more severe disease course and transcriptomic studies suggest that higher IFI-G expression in renal cells is associated with a poor response to conventional treatment. However further validation is required, and it remains unclear whether the poor outcomes in patients with high IFI-G expression in their kidneys are due to the direct effects of interferon on renal cells or indirectly through recruitment of inflammatory cells. In this study, we optimized a panel of antibodies for imaging mass cytometry (IMC) enabling examination of IFI-protein (P) expression and its association with immune cells infiltration and disruption of kidney architecture in archived renal biopsies for LN patients. Paraffin embedded renal biopsies from patients with LN who were part of the Lupus Nephritis New Emerging Team and University of Toronto Lupus Clinic cohorts were available for testing. We have previously demonstrated that IFI-P levels in the peripheral blood strongly correlate with IFI-G expression, suggesting that antibodies against IFI-Ps (ISG15, MX1, PKR) are reliable surrogates for gene expression. To facilitate standardization, a pseudo-tissue was created from peripheral blood that was composed of a mixture of IFN-stimulated and unstimulated peripheral blood mononuclear cells from a healthy control. This was subsequently mixed with plasma, clotted, and embedded in paraffin. This pseudo-tissue served as both a positive and negative control for IFI-P expression. In addition to renal biopsies, archived lymph node and tonsil tissues were stained to facilitate titration of antibodies directed against immune subsets. A panel of 25 metal-conjugated antibodies was successfully created that enabled staining renal resident cells/structure, infiltrating immune cells (including T cells, B cells, plasma cells, monocytes, macrophages, and dendritic cells), and IFI-Ps (Table 1). Testing in the pseudo-tissue confirmed that the antibodies directed against IFI-Ps (MX1, ISG15, PKR) effectively discriminated between IFN-stimulated and unstimulated cells, with a low staining background and that this tissue could be used for batch standardization for subsequent staining of a larger number of LN biopsies over time. In our preliminary studies, staining of 3 kidney biopsies revealed distinct patterns: with the minimal change biopsy showing no cellular infiltration and minimal expression of IFI-Ps; the membranous LN biopsy showing increased IFNAR levels but similar IFI-P expression to the minimal change biopsy and few immune cells; and the proliferative LN biopsy showing high levels IFI-Ps and many immune cells in the glomerulus and proximal tubules (Figure 1). Table 1. Validated IMC Panel Figure 1. Comparative IMC staining of adult renal biopsies with minimal change disease (A), class V LN (B) and class IV LN (C). The 3 IMC images for each sample represent images obtained from the same region and time, with a selection of glomerular, tubular and stromal markers shown in the upper panels, immune markers shown in the middle panels and IFI-P and IFNAR1 in the lower panels. MEG, megalin; VIM, vimentin; aSMA, a-smooth muscle actin; COL IV, collagen IV. This study validated an IMC approach for spatial analysis of interferon signatures and immune cell infiltration in LN kidney biopsies. Future analyses will apply this panel to a broader cohort of LN samples to identify biomarkers associated with treatment response. By enabling early stratification of LN patients, this approach could support the development of targeted therapies for individuals who are less likely to respond to standard treatments, ultimately improving long-term renal outcomes.

O022 / #466 Topic:AS14 - Innate Immunity ABSTRACT CONCURRENT SESSION 03: INNATE AND ADAPTIVE IMMUNITY IN SLE 22-05-2025 1:40 PM - 2:40 PM Systemic autoimmune rheumatic diseases (SARD) are a group of chronic diseases characterized by the presence of antinuclear antibodies (ANAs). However, ANAs cannot reliably be used as a diagnostic tool because a subset of healthy women are ANA+(~20%) and the majority of these individuals will not progress to SARD. Why some individuals progress while others remain asymptomatic is unknown. Previous work suggests that monocytes/DCs may support immunological disturbances observed in SARD, including a shift toward a T helper (Th) 17 cell phenotype with a concurrent decrease in Tregs. Our objective is to evaluate functional alterations in innate immune populations during SARD development. Experiments have been completed to examine the composition of innate immune cells in PBMCs using CITE-seq. Samples were used from 5 ANA-healthy controls, 11 ANA+asymptomatic (5 progressors sampled prior to progression, 6 nonprogressors) and 8 early SARD patients (4 SLE, 4 Sjogren’s disease). Five million freshly thawed PBMCs were depleted of T and B cells by negative selection, and stained with a panel of oligo-conjugated antibodies for the identification of DC/monocyte populations. 9000 cells were sequenced at a depth of 50000 reads for gene expression and 5000 reads for CITE-seq. Using both gene and surface protein expression, we identified 11 distinct DC and monocyte populations (Figure 1A). Proportional analysis revealed an expansion of nonclassical and activated nonclassical monocytes in progressor and SARD patients (Figure 1B). SARD patients also exhibited higher proportions of cDC3s, VCAN+ monocytes and MME+ DCs than ANA+individuals regardless of progression status (Figure 1B), suggesting a role for these cells in active disease. Comparing asymptomatic ANA+individuals, we found that classical, intermediate, and nonclassical monocytes were expanded in progressors while cDC1s, cDC2s and pDCs were expanded in nonprogressors (Figure 1B). Differential analysis showed high expression of interferon (IFN) stimulated genes in progressors and SARD patients (Figure 1C), implicating these proinflammatory cytokines in the transition to SARD. Interestingly, nonprogressors exclusively had elevated expression of heat shock proteins (Figure 1C), which have been shown to promote immunologic tolerance and may play a role in preventing progression to SARD. In contrast, progressors had elevated HLA expression (Figure 1C), suggesting enhanced antigen presentation capacity. These differences in gene expression were observed across cell types, but intermediate monocytes are shown as representative cells due to their role in antigen presentation. Several genes were differentially expressed between progressors and SARD (Figure 1C), potentially highlighting distinct roles for genes in initiating and driving disease. Further analysis will be done to examine pathways associated with these genes. Figure 1. A)UMAP of 11 annotated monocyte and DC clusters.B)Proportional analysis showing each cluster for control, non-progressor, progressor and SARD groups. Data was normalized to account for differences in cell counts per cluster and per patient sample.C)Heatmap showing differentially expressed genes between patient groups, with intermediate monocytes as a representative cell cluster (log2FC = |1.0|, Q > 0.05, min.pct = 0.1). Similar trends were seen in the remaining cell types. Genes and patients were clustered in an unsupervised manner. Genes of interest are indicated with pink representing genes elevated in progressors and green representing genes elevated in non-progressors. Our data reveals differences in proportions and gene expression between progressors, nonprogressors and SARD patients. Importantly, ANA+progressors show expanded monocyte populations and functional differences compared to nonprogressors prior to progression, highlighting immune disturbances even in the asymptomatic, preclinical stage of SARD. The results provide insight into the immune mechanisms that drive progression from asymptomatic autoimmunity to disease in SARD.

The presence of Anti-Nuclear Antibodies (ANA) are a hallmark of Systemic Autoimmune Rheumatic Disease (SARD) and can be present years before clinical diagnosis. Although this suggests that ANAs could be used as potential biomarker for disease progression, they are also found in up to 25% of women, only a small fraction of whom (<5%) will develop a SARD. As the onset of symptomatic disease is not infrequently associated with organ damage, there is tremendous interest in identifying biomarkers associated with a high risk of progression, which could potentially enable initiation of preventative therapy. Here, we sought to determine whether any of the tests typically available to rheumatologists can be used to identify ANA positive individuals at high risk of imminent progression. Study participants were recruited from the Early Autoimmune Rheumatic Disease Clinic at Toronto Western Hospital, where ANA positive (≥1:160 by IF or ≥1:80 with a specific autoantibody) individuals without a SARD diagnosis (based upon clinical classification criteria) were followed yearly, or earlier if they had new symptoms, for development of SARD symptoms. Participants either lacked SARD clinical criteria or had insufficient criteria for a SARD diagnosis (UCTD), with progression being defined as the onset of a new clinical criteria for SARD. All ANAs, complements, and specific autoantibodies were measured through the hospital laboratory, with specific ANAs being measured by Bioplex. 124 ANA positive individuals were followed by a minimum of 2 years, 15 of which clinically progressed within 2 years of follow-up. The mean age and proportion of female progressors did not differ significantly from that of non-progressors, however progressors were more likely to be non-Caucasian than non-progressors. Although the ANA titer and serum complement levels were similar in progressors and non-progressors, progressors had significantly more specific ANAs (1.53 ±1.41 vs 0.88 ± 0.92, p = 0.019, Student’s t test). With the exception of anti-dsDNA antibodies, all specific ANAs were more prevalent in progressors, but this difference was only significant for anti-La antibodies. The most prevalent antibody in the cohort in both progressors and non-progressors was anti-Ro (46.7% and 35.8%, respectively). Within anti-Ro positive individuals, discrimination between anti-Ro52 and -Ro60 showed that anti-Ro52 but not anti-Ro60 antibodies were significantly associated with progression, particularly when in tandem with a positive RF (Table 1). Table 1: Clinical and Serologic Associations with Progression The presence of anti-La and -Ro52 antibodies is associated with an increased risk of imminent clinical progression in the subsequent 2 years and such individuals merit close follow-up.

O048 / #626 Topic:AS15 - Lupus Nephritis-Clinical ABSTRACT CONCURRENT SESSION 08: RECENT ADVANCES IN LUPUS BIOMARKERS 23-05-2025 1:40 PM - 2:40 PM Lupus nephritis (LN) affects up to 50% of patients with lupus, of whom 40% will experience a subsequent renal flare, and up to 20% will progress to end-stage renal disease. Repeat kidney biopsies (KB) performed 2 years after the last LN flare have been shown to predict subsequent renal flares and renal dysfunction. In this study, we assessed whether 5 urinary biomarkers (UB), including CD163, MCP-1, Adiponectin, sVCAM-1 and PF4 measured 2 years after a LN flare, predict long-term renal outcomes. Patients who had a LN flare and stored urine 24±3 months after the LN flare were included in the study. The 5 UB levels were measured by ELISA 24±3 months after the LN flare. Examined renal outcomes: 1) Time to a subsequent LN flare (increase in proteinuria of at least 1000 mg/day if the baseline was <500 mg/day or doubling of proteinuria if the baseline was ≥500 mg/day, prompting a change in therapy) and 2) time to 30% decline in eGFR, after their 2-year urinary sample collection. 69 patients with LN were included. The median (IQR) follow-up time after their 2-year urinary sample collection was 129 (97.5-150) months. 50 patients achieved proteinuria of ≤700 mg at 2 years after the LN flare. This subcohort of patients had significantly lower UB levels 2 years after the LN flare compared to patients who persisted with proteinuria >700 mg (Figure 1). In this subcohort of patients, 27 (54%) experienced a subsequent LN flare with a median time to flare (IQR) of 3.5 (1.67-6.87) years, and 10 (20%) had a 30% decline in eGFR at a median time of 4.38 (3.73-5.33) years after their 2-year urinary sample collection. Elevated levels of MCP-1 (HR 1.13 (1.01-1.27), p=0.03) and CD163 (HR 1.48 (1.15-1.90), p=0.002) predicted a subsequent LN flare. While CD163 (HR 1.31 (1.10-1.57), p=0.002), Adiponectin (HR 1.53 (1.22-1.91), p=0.0002), sVCAM-1 (HR 1.11 (1.03-1.21), p=0.006), and PF4 (HR 1.14 (1.04-1.25), p=0.003) predicted a 30% decline in eGFR (Table 1). Figure 1. UB were significantly higher in patients who did not achieve an uPCR ≤700 mg (n=19) at 24±3 months after the LN flare as compared to those who did (n=50). Symbols represent the determination from a single individual, columns the median and the bars IQR. Table 1. Multivariable Cox Regression analysis. Predictors of adverse renal outcomes (Subcohort of patients who achieved a proteinuria of ≤700 mg at 24±3 months after the LN flare, N=50) UB measured 2 years after an LN flare predicted long-term renal outcomes.

BACKGROUND:Cognitive impairment (CI) is one of the most common manifestations of neuropsychiatric systemic lupus erythematosus (SLE). This study aimed to characterise the course of CI over a 1-year period in patients with SLE and its associated factors. METHODS:175 adult SLE patients from the University of Toronto Lupus Clinic were assessed at baseline, 6 months and 12 months using the American College of Rheumatology Neuropsychological Battery. CI was classified based on standardised z-scores in cognitive domains. Patients were categorised as persistent-CI (CI at all three time-points; T0, T1 and T2), never-CI (no CI at any time-point) or fluctuating-CI (CI at 1-2 assessments). Sociodemographic, clinical, laboratory and medication data were collected at each visit. Patients with persistent-CI were compared with never-CI patients. CI severity was determined based on the mean z-score of tests across all six domains. RESULTS:Over 1 year, 46% of patients experienced CI, with 17% showing persistent-CI, 29% fluctuating-CI and 54% never-CI. Persistent-CI patients exhibited more severe CI compared with fluctuating-CI. The most frequently affected cognitive domains were learning and memory, simple attention and processing speed, and visual-spatial construction. Factors associated with CI persistence over 1 year included Black race, older age at SLE diagnosis, divorced/separated status at T0 and higher disease-related damage at T0. CONCLUSION:This study highlights the variable nature of CI in SLE patients, with most exhibiting a stable course over 1 year. Factors such as sociodemographic characteristics and comorbidities may influence CI persistence.