Stage-resolved mapping of RP105 (CD180)-negative B cells reveals residual plasmablast-lineage skewing in SLE: a cross-sectional study Authors: Ayana Kanagawa, Akihito Maruyama, Tatsuki Shibuta, Mitsuteru Akahoshi, Sachiko Soejima, Yoshifumi Tada, Syuichi Koarada Published: 2026 ABSTRACT AbstractObjectiveTo determine whether clinically inactive/low-activity SLE reflects immunological remission in this cross-sectional study using CD180 (RP105)-anchored, stage-resolved profiling of peripheral B-cell differentiation.MethodsCD19+ B cells were analysed by flow cytometry in healthy controls (HCs; n=27), active SLE (n=29; sampled on admission before treatment initiation/escalation) and inactive/low-activity SLE (n=14; biological-free; ≥5 years of sustained Lupus Low Disease Activity State and/or Definition of Remission in SLE remission). Six ordered stages (subsets 0–5) were defined along the CD180-loss trajectory. CXCR5, CD11c, B-cell activating factor receptor and B-cell maturation antigen (BCMA) were quantified. Non-parametric tests with Benjamini-Hochberg false discovery rate (FDR) adjustment were applied.ResultsActive SLE redistributed towards downstream CD180− plasmablast-lineage stages (subsets 2–5) (14.07% vs 1.92% in HCs). Inactive/low-activity SLE retained a residual plasmablast-lineage (subset 2–5: 6.17%) expansion above HC levels despite durable quiescence. In HCs, CXCR5 positivity was high in upstream CD180+ compartments and declined stepwise across the trajectory. CXCR5 positivity at subset 0 was reduced in active SLE (FDR-adjusted p=0.0273) and inactive/low-activity SLE (FDR-adjusted p=0.0015) and at subset 3 in active SLE (FDR-adjusted p=0.0273) versus HCs. CD11c was enriched in upstream compartments during active disease, whereas BCMA was increased in mature downstream CD180− stages.ConclusionsCD180-anchored, stage-resolved profiling indicates that clinical quiescence may not fully equate to immunological normalisation in SLE. These cross-sectional observations do not establish durable persistence, and longitudinal validation is needed. An upstream CXCR5 decrement detectable even in long-term quiescence is consistent with extrafollicular-leaning priming changes arising before CD180 loss; however, any implications for longitudinal immunomonitoring remain hypothesis-generating. WHAT IS ALREADY KNOWN ON THIS TOPIC WHAT THIS STUDY ADDS Using a CD180/CD19/CD138-anchored, stage-resolved framework (six subsets), we show that inactive/low-activity SLE can retain an expanded CD180− plasmablast-lineage compartment compared with healthy controls. Using a CD180/CD19/CD138-anchored, stage-resolved framework (six subsets), we show that inactive/low-activity SLE can retain an expanded CD180− plasmablast-lineage compartment compared with healthy controls. CXCR5 is already reduced in the most upstream CD180+ compartment (subset 0) in both active and inactive/low-activity SLE, suggesting an early shift in follicular positioning potential. CXCR5 is already reduced in the most upstream CD180+ compartment (subset 0) in both active and inactive/low-activity SLE, suggesting an early shift in follicular positioning potential. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Stage-resolved immunophenotyping may help identify residual humoral activity despite clinical quiescence and generate hypotheses for longitudinal biomarker studies. Stage-resolved immunophenotyping may help identify residual humoral activity despite clinical quiescence and generate hypotheses for longitudinal biomarker studies. Introduction SLE is a prototypical systemic autoimmune disease driven in part by B-cell dysfunction, leading to pathogenic autoantibody production and inflammatory organ injury.1 Treat-to-target strategies and contemporary immunosuppression can achieve clinical remission or low disease activity; however, serological activity and humoral immune circuits may persist, raising the question of whether clinical quiescence truly reflects immunological remission.24 Circulating plasmablasts and related extrafollicular (EF) effector B-cell states expand during active SLE and can fluctuate rapidly depending on treatment timing, making single-population measurements difficult to interpret in routine practice.5 6 A practical, stage-resolved framework that captures the trajectory from upstream B cells to antibody-secreting lineages could therefore provide complementary biomarkers beyond conventional clinical indices. Recent immunological reviews emphasise that a naïve B cell’s decision to enter the germinal centre (GC) or the EF differentiation pathway is governed by the integration of activating and inhibitory signals during the first few days after antigen encounter.7 Early signalling through the PI3K/AKT/mTOR pathway—modulated by costimulatory molecules, Toll-like receptors (TLRs) and cytokines—helps determine whether B cells undergo GC maturation or rapidly differentiate into antibody-secreting cells via the EF route.7 Dysregulated PI3K/AKT/mTOR signalling, together with insufficient negative regulation, can bias these fate decisions towards EF responses and enhance the survival of autoreactive clones.7 Accordingly, mechanistic studies suggest that EF responses are not merely an accelerated version of GC maturation but instead reflect a distinct developmental programme, the overactivity of which may predispose to autoimmunity. The upstream triggers of EF differentiation are increasingly linked to innate receptor pathways and survival factors. CD180 (RP105), a TLR-homologous receptor expressed on most mature B cells, forms a heterodimer with MD-1 and negatively regulates TLR4 signalling; CD180 ligation attenuates TLR4-mediated responses.8 9 Conversely, loss of CD180 can cooperate with TLR7 and type I interferon (IFN) pathways to promote B-cell activation and antibody production, and CD180 ligation has been reported to inhibit IFN-α signalling through Lyn–PI3K–BTK-dependent mechanisms.10 11 CD180-negative B cells respond vigorously to B-cell activating factor (BAFF) and APRIL and preferentially express B-cell maturation antigen (BCMA) over BAFF receptor (BAFF-R), suggesting that survival cues further shape EF differentiation.1214 Downstream of these inputs, inhibitory receptors such as CD22 and FcγRIIB, together with FAS, constrain the PI3K/AKT/mTOR axis; defects in these checkpoints may tip the balance towards an EF fate.7 Taken together, these observations position CD180 as both a regulatory node and a phenotypic marker associated with EF-prone B-cell states. Importantly, these pathway-level inferences are increasingly supported by mechanistic studies that identify upstream drivers of EF commitment. In particular, receptor–ligand interactions and specialised T-cell help have been shown to actively promote EF differentiation in experimental models and are also suggested by patient-derived single-cell datasets. For example, in a TLR7-driven lupus model, B cells with reduced expression of the complement receptor CD21 were predisposed to differentiate into autoreactive EF effector cells, and CD21 downregulation was linked to engagement with endogenous ligands.15 Blocking CD21–ligand interactions mitigated CD21 loss and markedly reduced the generation of autoantibody-secreting cells.15 In parallel, single-cell analyses in paediatric SLE identified a Th10 population (a T helper subset with combined B-cell helper and cytotoxic features) that was expanded in lupus and was implicated in providing EF-type help to B cells within inflamed tissues.16 Regulatory T cells were also numerically abundant but could exhibit proinflammatory features, underscoring the complexity of the T–B interface in EF responses. Overall, these observations frame EF differentiation as a coordinated programme integrating innate receptor signalling, survival cues and specialised T-cell help, reinforcing the rationale for stage-resolved immunophenotyping beyond conventional serology. Loss of CD180 expression marks a lupus-associated B-cell compartment enriched for plasmablast-lineage features.9 17 18 Prior studies indicate that CD180-negative B cells respond to innate and IFN-related cues and contribute to immunoglobulin and autoantibody production, consistent with EF-leaning differentiation.19 20 Building on this biology, we developed an operational immunophenotyping approach anchored on CD180 status. Using CD180 and CD19 to partition peripheral CD19+ B cells, and CD138 to stage the CD180−CD19 low lineage, we delineated six phenotypically ordered subsets (0–5) spanning CD180+ upstream compartments through plasmablast-lineage and peripheral plasma cell-like stages. This framework was designed as an operational, stage-resolved phenotypic model rather than as direct proof of a strictly linear developmental trajectory. In this cross-sectional study, we compared healthy controls (HCs), active SLE and inactive/low-activity SLE (biological-free; ≥5 years of sustained Lupus Low Disease Activity State (LLDAS) and/or Definition of Remission in SLE (DORIS)-defined remission21 22) to evaluate whether CD180-negative subsets and EF-related phenotypes persist despite durable clinical quiescence, and to define stage-specific alterations in CXCR5, CD11c, BAFF-R and BCMA across these activity states. Participants were recruited from rheumatology services at International University of Health and Welfare and Saga University-affiliated institutions between May 2019 and June 2025. Patients in the inactive/low-activity group were evaluated clinically at least every 3–6 months and continuously fulfilled remission or low-disease-activity criteria over the 5 years preceding sampling. The 5-year threshold was selected pragmatically to enrich for sustained clinical quiescence, minimise misclassification and reduce the influence of recent flare-related or treatment-related immunological fluctuations. We acknowledge that other thresholds could also be justified and that this choice may limit generalisability. Methods Study design, ethics and participants We conducted a cross-sectional flow cytometry study of peripheral blood CD19+ B cells from patients with SLE and HCs. SLE was classified according to the 2019 European Alliance of Associations for Rheumatology (EULAR)/American College of Rheumatology (ACR) criteria and confirmed by the treating rheumatologist. Eligible participants were aged ≥18 years and had an available peripheral blood sample. Exclusion criteria included active infection at sampling or pregnancy. HCs had no history of autoimmune disease or chronic infection and were not receiving immunomodulatory medication. No participants had received B-cell-depleting therapy. Trial registration: Not applicable. Patient and public involvement: Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research. Definition of clinical groups The active SLE group comprised patients sampled at hospital admission before treatment initiation or intensification. This group included both newly diagnosed, treatment-naïve cases and flare cases that had previously been clinically inactive/low activity. To minimise treatment-related confounding, the inactive/low-activity group was restricted to patients who were not receiving biological therapy (including belimumab and rituximab) and who had maintained continuous clinical inactivity/low-activity for ≥5 years immediately prior to sampling, as defined by LLDAS and/or DORIS remission.21 22 Patients in this group were evaluated clinically at least every 3–6 months and continuously fulfilled remission or low-disease-activity criteria over the 5 years preceding sampling. The 5-year threshold was selected pragmatically to enrich for sustained clinical quiescence, minimise misclassification and reduce the influence of recent flare-related or treatment-related immunological fluctuations; we acknowledge that other thresholds could also be justified and that this choice may limit generalisability. Sample collection and PBMC preparation Peripheral blood was collected into EDTA tubes at the time of clinical evaluation. For the active SLE group, samples were obtained at hospital admission prior to treatment initiation or escalation. For the inactive/low-activity SLE group, samples were obtained during routine outpatient visits. HCs were recruited as volunteers from the university and affiliated hospitals. Peripheral blood mononuclear cells (PBMCs) were isolated as soon as possible after blood collection (typically within 2 hours) by Ficoll density-gradient centrifugation and analysed fresh. Antibody staining and flow cytometry acquisition PBMCs were stained with a standardised multicolour antibody panel (online supplemental table 1) and acquired on a BD FACSAria II flow cytometer. Compensation was performed using single-stained controls. All samples were acquired using harmonised instrument settings and analysed with a pre-specified gating template. Positivity gates were defined using fluorescence-minus-one (FMO) controls and applied consistently across batches, including the thresholds used to delineate double-negative and double-positive populations for each marker. It should be noted that our phenotyping panel did not include CD27; therefore, plasmablast-lineage subsets were operationally defined on the basis of CD180, CD19 and CD138 expression, and these designations should be interpreted within that context. Gating strategy and stage-resolved subset definition A prespecified sequential gating strategy (figure 1A,B) defined six phenotypically ordered subsets (0–5) within CD19+ B cells based on CD180 and CD19, with CD138 used to stage the CD180−CD19low lineage. The sequential gating strategy was defined as follows. Subset 0 comprised CD19-positive cells with high RP105 expression, whereas subset 1 comprised CD19-positive cells with low RP105 expression. Subsets 2–5 were defined within the RP105-negative compartment. Among these, subset 2 was characterised by relatively high CD19 expression. The remaining RP105-negative CD19-low intermediate gate (subset A) was further subdivided according to CD138 expression into subset 3 (CD138-negative), subset 4 (CD138-low) and subset 5 (CD138-high). Positivity thresholds were established using FMO controls and applied consistently across batches. These subsets were used as an operational, stage-resolved framework for comparative immunophenotyping across clinical states and were not intended to imply direct lineage tracing or a strictly unidirectional maturation pathway. Operational definition of CD180 (RP105)-negative B-cell stages. (A) Representative sequential gating strategy using CD180 (RP105), CD19 and CD138 to define subsets 0–5 within peripheral blood CD19+ B cells from a patient with active SLE. Subset A (Gate A) denotes the intermediate CD19low CD180− gate within which subsets 3–5 were defined. (B) Schematic overview of the six ordered stages (subsets 0–5) and their defining features, spanning CD180hi and CD180low upstream compartments and a CD180− plasmablast-lineage continuum. CXCR5 was quantified as a marker of follicular positioning potential; reduced CXCR5 was interpreted as diminished follicular homing potential, consistent with EF-leaning differentiation.6 7 BAFF-R and BCMA were quantified as stage-related receptor readouts within the CD180-anchored framework, and CD11c was quantified as an upstream priming marker.23 The BAFF-R/BCMA ratio was calculated solely as a descriptive summary to aid visualisation of relative receptor balance across subsets and was not subjected to inferential testing or multiplicity correction. Accordingly, this ratio was not treated as an independent biological or mechanistic endpoint. Statistical analysis Analyses were prespecified as subset-resolved comparisons (subsets 0–5) across clinical states. Overall group differences (HCs, active SLE, inactive/low-activity SLE) were tested using Kruskal-Wallis tests, followed by prespecified two-sided Mann-Whitney U tests for three pairwise group contrasts (HCs vs active SLE, HCs vs inactive/low-activity SLE and active vs inactive/low-activity SLE). For the four marker-based analyses (CXCR5, CD11c, BAFF-R and BCMA), these comparisons were evaluated across six subsets, yielding 72 prespecified pairwise comparisons in total (4 markers×3 contrasts×6 subsets). Benjamini-Hochberg false discovery rate (BH-FDR) correction was applied within each marker and within each pairwise contrast across the six subsets, rather than globally across all markers, yielding adjusted p values (p_adj). We selected subset-level Mann-Whitney U tests because the primary aim was to compare biologically predefined stages individually within the CD180-anchored framework, and because the modest sample size, non-normal distributions and potentially non-linear differences across subsets made more complex ordered-factor models less robust for the present exploratory cross-sectional dataset. Although approaches incorporating subset as an ordered factor, including mixed-effects modelling, may be valuable in larger or longitudinal datasets, we considered such models potentially overparameterised for the current study design. Non-parametric analyses were performed in IBM SPSS Statistics (V.27.0.1.0); BH-FDR adjustment was implemented using a prespecified spreadsheet and verified in R (V.4.4.3). Data are summarised as mean (SD) unless otherwise specified; missing data were not imputed. Major therapies were summarised by group (online supplemental table 2). Gating and marker quantification were performed using coded samples with group labels concealed until completion of the primary analysis. No formal a priori sample-size calculation was performed; all eligible participants with available samples were included. Despite BH-FDR correction, residual false-positive risk remains given correlated markers and multiple subset-level comparisons; findings should be interpreted cautiously and validated in independent and longitudinal cohorts. Results Stage-resolved quantification reveals redistribution and incomplete reversion towards healthy CD180 (RP105)− B-cell state distributions across SLE activity strata. Using a gating strategy based on CD180, CD19 and CD138 (figure 1A,B), we quantified the distribution of subsets (0–5) among peripheral blood CD19+ B cells in HCs (n=27), active SLE (n=29) and inactive/low-activity SLE (n=14; DORIS-defined remission n=4, LLDAS n=10). In HCs, the CD180− compartment (subsets 2–5) was rare (1.92±1.60% of CD19+ B cells), with most cells residing in CD180+ upstream subsets (subsets 0–1: 98.08±1.60%) (figure 2B). In active SLE, the B-cell landscape shifted towards CD180− states (subsets 2–5: 14.07±10.80%), exceeding HC levels (p_adj=2.05×10−6). This increase was primarily driven by accumulation of plasmablast-lineage cells (subset 3: 8.73±7.28%; p_adj=1.07×10−6 vs HCs). Stage-resolved redistribution of subsets 0–5 across clinical states. (A) Representative pie charts showing the subset composition (subsets 0–5) within peripheral CD19+ B cells in an HC, a patient with active SLE and a patient with inactive/low-activity SLE. (B) Within-group subset frequencies (subsets 0–5) shown as bar plots (mean±SD; error bars indicate SD). (C) Between-group comparisons for each subset. Prespecified pairwise contrasts (HC vs active SLE, HC vs inactive/low-activity SLE and active vs inactive/low-activity SLE) were tested using two-sided Mann-Whitney U tests. P values were adjusted using the BH-FDR procedure within each contrast across the six subsets, yielding BH-FDR-adjusted p values (p_adj). *p_adj<0.05, **p_adj<0.01, ***p_adj<0.001. BH-FDR, Benjamini-Hochberg false discovery rate; HCs, healthy controls. Inactive/low-activity SLE also showed an expanded CD180− compartment (6.17±5.93%) compared with HCs (p_adj=0.0103), despite meeting LLDAS/remission criteria. Subset 3 likewise remained increased (3.76±4.67%; p_adj=0.0162 vs HCs), indicating that clinical inactivity can coexist with a residual plasmablast-skewed B-cell compartment (figure 2B,C). Compared with active SLE, inactive/low-activity SLE showed significant reductions in both the total CD180− compartment (subsets 2–5; p_adj=0.0213) and subset 3 (p_adj=0.0107), consistent with graded attenuation across the activity spectrum. Thus, although overall subset distributions in inactive/low-activity SLE shifted towards the healthy pattern, CD180− plasmablast-lineage cells persisted above healthy levels. Group-level summary statistics are provided in online supplemental table 3. Stage-resolved CXCR5 profiling suggests stepwise loss along the CD180 (RP105)− route and an upstream decrement that persists despite clinical quiescence. We next examined CXCR5, a chemokine receptor that reflects follicular positioning potential, across phenotypically defined stages (figure 3). In HCs, CXCR5 positivity was high in upstream CD180+ compartments (subsets 0–1) and declined stepwise across downstream CD180− subsets, approaching minimal levels by subsets 4–5. This along-trajectory decline is consistent with progressively diminished follicular positioning potential as cells transition towards the plasmablast lineage along the CD180− route. Stage-resolved CXCR5 expression across the CD180-defined trajectory. (A) Representative histograms of CXCR5 expression by subset. (B) Within-group CXCR5 positivity (% positive) across subsets shown as bar plots (mean±SD; error bars indicate SD). Positivity denotes the proportion of CXCR5-positive events within each phenotypically defined subset. (C) Between-group comparisons for each subset. Prespecified pairwise contrasts (HC vs active SLE, HC vs inactive/low-activity SLE and active vs inactive/low-activity SLE) were tested using two-sided Mann-Whitney U tests. P values were adjusted using the BH-FDR procedure within each contrast across the six subsets, yielding BH-FDR-adjusted p values (p_adj). *p_adj<0.05, **p_adj<0.01, ***p_adj<0.001. BH-FDR, Benjamini-Hochberg false discovery rate; HCs, healthy controls. In active SLE, CXCR5 positivity was already reduced at the earliest stage (subset 0) compared with HCs (p_adj=0.0273), consistent with the left-shift of the corresponding histogram (figure 3A). A downstream difference was also observed at the plasmablast stage (subset 3; p_adj=0.0273 vs HCs). Subset 2 showed a directionally similar trend (p_adj=0.0624 vs HCs) that did not remain significant after correction. Importantly, in inactive/low-activity SLE, CXCR5 positivity at subset 0 was likewise significantly lower than in HCs after BH-FDR correction (p_adj=0.0015). By contrast, across downstream CD180− stages, CXCR5 positivity was numerically intermediate between active SLE and HCs, but no downstream comparisons remained significant after correction (eg, subset 3: p_adj=0.358 vs HCs). Active versus inactive/low-activity contrasts were also not significant after correction (subset 0: p_adj=0.752; subset 3: p_adj=0.210). Collectively, these data show a stepwise decline in CXCR5 along the CD180-defined phenotypic trajectory, identifying an upstream CXCR5 decrement that remains detectable even in long-term clinical quiescence. Stage-resolved BAFF-R and BCMA profiling demonstrates stage-dependent remodelling of survival-receptor patterns along the CD180 (RP105)− trajectory. We assessed BAFF-R and BCMA across subsets (0–5) to interrogate survival-receptor patterns along the CD180− differentiation route (figure 4). In HCs, BAFF-R was broadly expressed in upstream compartments. Across clinical groups, BAFF-R showed no significant between-group differences at any subset after BH-FDR correction. Stage-resolved BAFF-R and BCMA expression and descriptive visualisation of the BAFF-R/BCMA balance. (A) BAFF-R expression (MFI, arbitrary units) across subsets shown as bar plots (mean±SD; error bars indicate SD). (B) BCMA expression (MFI, arbitrary units) across subsets shown as bar plots (mean±SD; error bars indicate SD). Where indicated, prespecified pairwise contrasts were tested using two-sided Mann-Whitney U tests. P values were adjusted using the BH-FDR procedure within each contrast across the six subsets, yielding BH-FDR-adjusted p values (p_adj). *p_adj<0.05, **p_adj<0.01, ***p_adj<0.001. (C) BAFF-R/BCMA balance expressed as the ratio of group-level subset medians (BAFF-R median divided by BCMA median for each subset) and plotted on a logarithmic scale (descriptive visualisation only; ratios were not subjected to statistical testing). BAFF-R, B-cell activating factor receptor; BCMA, B-cell maturation antigen; BH-FDR, Benjamini-Hochberg false discovery rate; HCs, healthy controls; MFI, mean fluorescence intensity. In contrast, BCMA mean fluorescence intensity (MFI) differed among groups in downstream stages (BH-FDR-adjusted Kruskal-Wallis p_adj=0.0309 for subset 3, 0.0361 for subset 4 and 0.0309 for subset 5). Pairwise testing indicated that active SLE exhibited higher BCMA in mature CD180− compartments (subsets 4–5) compared with HCs (subset 4: p_adj=0.0048; subset 5: p_adj=0.0016). In inactive/low-activity SLE, BCMA levels shifted towards the healthy pattern overall and were lower than in active SLE at the plasmablast stage (subset 3; p_adj=0.0242), consistent with attenuation of this survival-axis feature with clinical control. For visualisation only, we summarised BAFF-R/BCMA balance per subset as a descriptive overview of relative receptor patterns across stages (figure 4C). Although active SLE showed numerically lower BAFF-R/BCMA ratios across CD180− stages (subsets 2–5), this readout was not subjected to statistical testing and should not be interpreted as an inferential endpoint. Given the substantial MFI variability and the absence of robust between-group differences in BAFF-R after correction, the ratio is presented only as a descriptive visual summary. CD11c enrichment is most evident in upstream CD180 (RP105)-defined compartments during active SLE and is not sustained after BH-FDR correction in inactive/low-activity SLE. CD11c is an activation-associated marker frequently linked to atypical/EF-leaning B-cell phenotypes. We quantified the frequency of CD11c+ cells across subsets (0–5) (figure 5). In HCs, CD11c positivity was low across stages. In active SLE, CD11c+ frequencies were increased in upstream subsets (0–1), with significant elevations versus HCs in subset 0 and subset 1 (both p_adj=0.0202). In downstream CD180− stages (subsets 2–5), CD11c+ frequencies were numerically higher in active SLE than in HCs, but these differences did not remain significant after BH-FDR correction. CD11c expression highlights an upstream priming feature within the CD180-defined framework. (A) Representative histograms of CD11c expression by subset. (B) Within-group CD11c positivity (% positive) across subsets shown as bar plots (mean±SD; error bars indicate SD). (C) Between-group comparisons for each subset. Prespecified pairwise contrasts (HC vs active SLE, HC vs inactive/low-activity SLE and active vs inactive/low-activity SLE) were tested using two-sided Mann-Whitney U tests. P values were adjusted using the BH-FDR procedure within each contrast across the six subsets, yielding BH-FDR-adjusted p values (p_adj). *p_adj<0.05, **p_adj<0.01, ***p_adj<0.001. BH-FDR, Benjamini-Hochberg false discovery rate; HC, healthy control. In inactive/low-activity SLE, CD11c+ frequencies in upstream subsets were numerically intermediate between HCs and active SLE; however, neither inactive versus HC nor active versus inactive contrasts reached BH-FDR significance at any stage. Collectively, these data show that CD11c enrichment is concentrated in early CD180-defined compartments during active disease, whereas any residual upstream elevation in inactive/low-activity SLE is modest and not supported after multiple-testing correction (figure 5A–C). Integrated summary across stage-resolved readouts Together, stage-resolved profiling across CD180-defined subsets revealed activity-stratified redistribution of B-cell states, characterised by expansion of the CD180− plasmablast-lineage compartment in active SLE and a residual increase above healthy levels in inactive/low-activity disease (figures 2–5). Across the phenotypic ordering, CXCR5, CD11c and BCMA displayed stage-dependent patterns that varied by clinical state. Effect-size estimates were concordant with the main p value-based findings (online supplemental table 4). In particular, residual CD180− expansion (subsets 2–5), plasmablast-stage frequency (subset 3) and CXCR5 reduction at subset 0 showed moderate-to-large rank-biserial correlations in the key group comparisons. By contrast, CD11c showed more moderate and variable patterns, and BCMA was more heterogeneous, supporting more cautious interpretation of these markers than of the principal subset and CXCR5 comparisons. Overall, these readouts provide a stage-resolved descriptive framework for future biomarker studies rather than evidence of clinically ready immunomonitoring. Exploratory analyses within the inactive/low-activity cohort did not identify clear associations between residual CD180− expansion and disease duration. Specifically, disease duration was not significantly correlated with the frequency of subsets 2–5 (Spearman’s rho=0.099, p=0.736), subset 3 (rho=0.119, p=0.686) or subsets 3–5 (rho=0.088, p=0.765). Individual-level distributions of subsets 0–5 showed interindividual variability (online supplemental figure 1). Subset 3 frequency was also not significantly correlated with anti-double-stranded DNA (anti-dsDNA) levels (Spearman’s rho=0.233, p=0.423; online supplemental figure 2). Discussion These results provide stage-resolved evidence that clinical quiescence may not fully equate to immunological normalisation in SLE. The observed between-group differences are consistent with residual EF-leaning immunological skewing during clinical quiescence, but the present cross-sectional data do not establish temporal persistence or within-individual trajectories. Alternative explanations, including cohort enrichment, unmeasured low-grade disease activity, redistribution of relevant cell populations, treatment-related effects and survivor bias, should also be considered. Our results corroborate that SLE is characterised by an expanded CD180−, plasmablast-skewed B-cell compartment during active disease, consistent with prior work implicating EF responses in lupus.5 6 18 19 23 24 In clinically inactive/low-activity SLE, this compartment remained expanded above healthy-control levels, although to a lesser degree than during flare, supporting the view that clinical quiescence may correspond to only partial immunological normalisation. At the same time, the cross-sectional design does not allow us to conclude that this residual skewing reflects durable within-individual persistence. The subset structure used here also requires careful interpretation. The ordering of subsets 0–5 along the CD180-loss trajectory is biologically plausible and useful for stage-resolved comparison, but the present data do not directly establish a strictly linear maturation pathway. Alternative or partially overlapping differentiation routes may generate similar phenotypic distributions. Thus, the framework should be understood as an operational, CD180-anchored phenotypic ordering rather than definitive lineage tracing, and CD180 loss should not be interpreted as a strictly irreversible or unidirectional differentiation marker. In parallel, CD11c enrichment was concentrated in upstream CD180-defined compartments during active SLE, pointing to early EF-leaning activation signatures that may precede or accompany plasmablast expansion. Downstream, BCMA was increased in mature CD180⁻ stages in active disease, consistent with stage-related survival-receptor remodelling accompanying plasmablast-lineage differentiation. However, the BAFF-R/BCMA balance should not be overinterpreted. Although it may provide a convenient descriptive overview of receptor patterns across subsets, BAFF-R differences were not robust after multiple-testing correction and MFI variability was substantial. We therefore regard this ratio as a descriptive visual summary rather than independent mechanistic or inferential evidence. An important feature of the present study was the deliberate enrichment of patients with prolonged clinical quiescence. The inactive/low-activity group was restricted to biological-free patients who had maintained sustained LLDAS and/or DORIS-defined remission for at least 5 years before sampling. This threshold was selected pragmatically to enrich for long-term quiescence, minimise misclassification and reduce the influence of recent flare-related or treatment-related immunological fluctuations. Patients in this group were clinically assessed at least every 3–6 months during that interval. This design strengthens internal consistency and helps reduce the influence of transient peri-flare changes and recent treatment-related fluctuations. However, it does not eliminate confounding. Background therapy differed across groups, and long-term immunosuppressive treatment may itself influence subset composition and survival-receptor expression, including BAFF-R/BCMA-related patterns. Historical hydroxychloroquine exposure may also have differed, particularly because hydroxychloroquine became available in Japan only in 2015. Exploratory treatment-stratified analyses within the inactive/low-activity cohort did not identify a clear association between current immunosuppressant use and residual CD180− expansion or BAFF-R/BCMA-related readouts; however, these comparisons were markedly limited by the very small number of non-users and by missing receptor-level data in some cases, and should therefore be interpreted cautiously. Treatment-related effects cannot be excluded in the current dataset. Therefore, although the residual CD180− plasmablast-lineage bias observed in clinically quiescent patients is notable, we cannot fully separate disease biology from treatment-related effects in the current dataset. Because the active cohort was sampled at admission prior to treatment initiation or escalation, whereas inactive/low-activity patients were sampled during routine follow-up, our primary inference concerns the presence of a residual plasmablast-lineage footprint and associated stage-resolved signatures in clinically quiescent patients rather than causal effects of contemporaneous therapy. Accordingly, these findings are descriptive and should not be interpreted as causal. These findings provide a stage-resolved descriptive framework for future biomarker and mechanistic studies. Compared with single-population plasmablast measurements, this approach may better capture broader shifts across the B-cell differentiation landscape and may help test whether upstream changes precede or accompany downstream CD180− plasmablast-lineage expansion. However, implications for flare prediction, stage-guided immunomonitoring, precision immunotherapy, CAR-T or BCMA-targeting strategies remain hypothesis-generating and require prospective longitudinal validation. Exploratory analyses within the inactive/low-activity cohort did not identify clear correlations between residual CD180− expansion, anti-dsDNA titres or disease duration, but these analyses were underpowered and should be interpreted cautiously. Several limitations should be considered. First, the cross-sectional design precludes within-individual inference across flare-to-quiescence transitions; longitudinal studies are required to determine whether upstream EF-leaning features precede downstream changes and predict subsequent flares. Second, our compact phenotyping panel did not include CD27; therefore, the plasmablast-lineage subsets in this study were defined operationally on the basis of CD180, CD19 and CD138 rather than by a conventional CD27/CD38 plasmablast definition. Third, although the inactive/low-activity group was biological-free at sampling and met stringent criteria of sustained LLDAS and/or DORIS remission for more than 5 years, background therapy differed across groups and residual confounding by treatment history and sampling context cannot be excluded. Disease duration also differed between the SLE groups, and the active cohort included newly diagnosed cases. HCs were younger than the SLE groups, and age-related differences in B-cell phenotypes cannot be excluded. Finally, although BH-FDR correction was applied to control false discovery, residual false-positive risk remains given multiple correlated markers and subset-resolved comparisons; therefore, findings should be interpreted cautiously and validated in independent and longitudinal cohorts. Despite these caveats, our data support a coherent framework in which EF-leaning priming is detectable in upstream compartments, loss of CD180 aligns with downstream plasmablast-lineage differentiation and stage-resolved readouts capture residual immunological activity despite clinical control. In conclusion, stage-resolved mapping through the lens of CD180 loss reveals a residual plasmablast-lineage footprint in clinically inactive/low-activity SLE, indicating that clinical quiescence may not fully equate to immunological normalisation in SLE. These cross-sectional observations do not establish durable persistence, and longitudinal validation is needed. Supplementary material