Anolik2004 - Rituximab and B Cell Abnormalities in SLE

Full citation: Anolik JH, Barnard J, Cappione A, Pugh-Bernard AE, Felgar RE, Looney RJ, Sanz I. Rituximab improves peripheral B cell abnormalities in human systemic lupus erythematosus. Arthritis Rheum. 2004;50(11):3580–3590. doi:10.1002/art.20592

Raw file: [[raw/Anolik2004.pdf]]

Summary

This study is the first mechanistic examination of B cell subset dynamics before, during, and after selective CD20-targeted B cell depletion in human SLE. As part of a phase I/II dose-escalation trial of rituximab (n=17 analyzable patients; low, intermediate, and high-dose arms), the authors profiled peripheral blood B cell subsets by flow cytometry using IgD/CD27 four-quadrant gating and CD38/IgD Bm1–Bm5 classification, and tracked autoreactive VH4.34 B cells by 9G4 antiidiotype staining. Baseline phenotyping confirmed three characteristic SLE abnormalities: naive lymphopenia, double-negative (IgD⁻CD27⁻) B cell expansion, and plasmablast expansion. Seven normal controls provided reference values.

After effective B cell depletion and immune reconstitution (≥1 year post-treatment), naive lymphopenia, DN expansion, and plasmablast expansion all resolved significantly — establishing that these abnormalities are driven by ongoing B cell dysregulation rather than irreversible programming. Autoreactive VH4.34 memory B cells also normalized. In contrast, pre-GC (IgD⁺CD38^high) expansion persisted or worsened after rituximab in the 3 patients with high baseline pre-GC frequencies, suggesting either intrinsic resistance of this population or ongoing GC activity that replenishes it.

Anti-dsDNA serum autoantibodies remained elevated in most patients despite B cell normalization, raising mechanistic questions about the relative contribution of long-lived plasma cells versus memory B cell–derived short-lived plasmablasts to sustained autoantibody production. The authors propose a two-component model: a short-lived plasmablast pool (indirectly depleted by rituximab within months) and a long-lived plasma cell component (resistant to depletion, residing in bone marrow or inflamed tissue niches).

Study Design

• Type: Phase I/II dose-escalation interventional trial with longitudinal B cell subset monitoring
• Sample size: 17 analyzable patients (19 enrolled; 2 withdrew); 7 normal controls for baseline comparison; subset of 11 patients with IgD/CD27 phenotyping at baseline
• Setting: University of Rochester Medical Center; active adult SLE (SLAM score >5); follow-up at 1, 2, 3, 6, 9, 12 months post-infusion (some up to 36 months)
• Population: Adults ≥18 years, ACR-criteria SLE, clinically active; mostly female; Caucasian/Black/Hispanic/Asian; excluded cyclophosphamide or high-dose prednisone; most on background azathioprine, hydroxychloroquine, or mycophenolate
• Rituximab dosing: Low dose (1 × 100 mg/m²), intermediate dose (1 × 375 mg/m²), high dose (4 × 375 mg/m² weekly)
• Effective depletion defined: CD19⁺ B cells <1% of total PBL, or absolute count <5 cells/µl (achieved in 11 of 17)

Key Findings

• Naive lymphopenia: SLE 35 ± 17% vs. controls 68 ± 6% (IgD⁺CD27⁻; P=0.0008, n=8 SLE, n=7 controls); normalized after effective depletion + reconstitution (P=0.008 for effective depletors).
• DN expansion at baseline: SLE 14.4 ± 7.9% vs. controls 3.9 ± 1.9% (IgD⁻CD27⁻; P=0.01, n=8 SLE, n=7 controls); correlates with VH4.34 autoantibody titers (R²=0.8, P<0.05); normalizes after effective depletion (P=0.05 vs. baseline).
• Plasmablast expansion at baseline: SLE 18.5 ± 17.9% (range 0–51%) vs. controls 0.24 ± 0.23% (range 0–0.5%; P=0.001, n=15 SLE); decreases to 2.4% after effective depletion + reconstitution (P=0.009).
• Plasmablast identification: CD38^high, CD19^low, CD20⁻ or CD38^high, IgD⁻, CD20⁻; confirmed as distinct from pre-GC cells (CD38^high, CD19⁺, CD20⁺, CD10⁺).
• Rapid plasmablast decline despite CD20⁻ phenotype: In select patients (e.g., patient 11), plasmablast frequency dropped from 40% at baseline to 14% at 2 months post-infusion — despite rituximab not directly targeting CD20⁻ cells. This implies a short-lived population continuously replenished by the CD20⁺ B cell pool.
• Pre-GC (Bm2′) expansion: CD38^high, CD19⁺, CD20⁺, CD10⁺ (IgD⁺) cells were expanded in only 3 of 15 patients at baseline (mean 4.7 ± 2.8% in controls vs. 9 ± 11.9% overall); notably, all 3 patients with marked pre-GC expansion had incomplete or transient rituximab depletion, suggesting these cells are either rituximab-resistant or mark a biologically distinct patient subset.
• Residual B cells after effective depletion: Predominantly switched memory phenotype (CD20⁺, CD38^low, CD27⁺, IgD⁻), with a minor CD20⁻, CD38^high plasma cell fraction; absolute memory B cell counts dramatically reduced (0.21 vs. 76.8 cells/µl at baseline).
• Autoreactive VH4.34 memory B cells: 16.2 ± 11.9% in SLE at baseline vs. 1.3 ± 0.3% in healthy controls (P=0.03, n=6 SLE); returns to 1.92 ± 0.7% after effective rituximab depletion (n=9 depletors). Interpreted as reflecting restoration of GC tolerance checkpoints.
• Anti-dsDNA persistence: Serum anti-dsDNA did not decrease significantly by 1 year overall (P=0.3); only 4 of 8 evaluable patients had serologic improvement; normalization in some patients required 24–36 months.
• B cell depletion variability: 6 of 17 patients failed effective depletion; FcγRIIIa genotype and serum rituximab levels were previously reported as determinants; B cell activation markers at baseline (high pre-GC or plasmablast frequency) were associated with incomplete depletion.

Methods Used

• Conventional Flow Cytometry — FACSCalibur; 5-color maximum panels; IgD/CD27 four-quadrant gating; CD38/IgD Bm1–Bm5 classification; CD19/CD38 plasmablast identification
• Bm Classification — Bm1–Bm5 IgD/CD38 staging applied as primary subset resolution tool alongside IgD/CD27 gating
• Ficoll-Hypaque PBMC isolation; CD19 magnetic selection for residual B cell enrichment

Entities Mentioned

Double-Negative B Cell, Plasmablast, CD19, CD20, CD27, CD38, IgD, IgG

Concepts Addressed

Extrafollicular Response, Germinal Center, Memory B Cell, Class Switch Recombination, Somatic Hypermutation

Relevance & Notes

This paper is from the same Rochester/Sanz laboratory as Wei2007 - DN Memory B Cells in SLE, published three years earlier. It provides the first systematic quantification of DN B cell and plasmablast expansion in SLE with matched controls, using the same IgD/CD27 gating framework that Wei2007 subsequently applied to characterise DN cell identity. Reading them together, Anolik2004 establishes the clinical phenotype and demonstrates reversibility; Wei2007 characterises the underlying biology (SHM, functional responses, FcRH4 expression).

For the dengue wiki, the key contribution is a mechanistic model for acute plasmablast dynamics: circulating CD20⁻ plasmablasts in SLE are short-lived and continuously replenished by CD20⁺ precursors, demonstrated by their rapid decline after rituximab. If dengue plasmablasts (which also arise as CD20⁻ CD38^high cells) follow the same biology, the massive acute-phase expansion in dengue days 7–10 represents a wave of EF differentiation from the naive/memory pool, not long-lived cell accumulation — with implications for the durability and quality of the resulting antibody response.

Limitations: Small n per dose group; dose-escalation design complicates within-group comparisons; subset phenotyping was done only in a subset of patients at baseline (n=8 for IgD/CD27, n=11 for broader phenotyping); the Bm2ʹ pre-GC population defined by CD38^high/IgD⁺/CD20⁺ was not universally validated with CD10 at all time points.

Questions Raised

• Are circulating plasmablasts in acute dengue similarly short-lived and CD20⁻? If so, how do dengue plasmablast kinetics (peak days 7–10, rapid decline) compare to the rituximab-induced depletion kinetics observed here?
• Does the Bm2ʹ pre-GC cell population expand in acute dengue, and does it track with eventual seroconversion or GC formation? It was not tracked in most early dengue flow cytometry papers.
• Rituximab normalises VH4.34 autoreactive memory B cells — interpreted as restoration of GC censoring. Does dengue B cell dysregulation involve similar failures of GC tolerance, contributing to cross-reactive or autoreactive antibodies that associate with severe disease?