Common connective tissue disorder and anti-cytokine autoantibodies are enriched in idiopathic multicentric castleman disease patients

Abstract

Introduction: Idiopathic Multicentric Castleman Disease (iMCD) is a polyclonal lymphoproliferative disorder involving cytokine storms that can lead to organ failure and death. The cause of iMCD is unknown, but some clinical evidence suggests an autoimmune etiology. For example, connective tissue disorders (CTDs) and iMCD share many clinical features, and autoantibodies have been anecdotally reported in individual iMCD patients. This study investigates whether common autoantibodies are shared across iMCD patients.

Methods: We assembled custom bead-based protein arrays consisting of 52 autoantigens traditionally associated with CTDs and 38 full-length cytokines and screened serum samples from 101 iMCD patients for IgG autoantibodies. We also screened samples with a 1,103-plex array of recombinant human protein fragments to identify additional autoantibody targets. Finally, we performed receptor blocking assays on select samples with anti-cytokine autoantibodies (ACAs) identified by array.

Results: We found that an increased proportion of iMCD patients (47%) tested positive for at least one CTD-associated autoantibody compared to healthy controls (HC) (17%). Commonly detected CTD-associated autoantibodies were associated with myositis and overlap syndromes as well as systemic lupus erythematosus (SLE) and Sjögren's Syndrome (SS). ACAs were also detected in a greater proportion of iMCD patients (38%) compared to HC (10%), while the protein fragment array identified a variety of other autoantibody targets. One iMCD sample tested positive for receptor blocking against interferon-ω (IFNω).

Discussion: IgG autoantibodies binding autoantigens associated with common CTDs and cytokines are elevated in iMCD patients compared to HC, suggesting that autoimmunity may be involved in iMCD pathogenesis.

Keywords: TAFRO; autoantibody; autoimmunity; connective tissue disorders; iMCD; luminex; protein array.

Figure 1

Autoantibodies associated with CTDs are prevalent in iMCD patients. (A) Heatmap displaying serum IgG AAbs identified by a 52-plex array of autoantigens associated with traditional CTDs in iMCD patients (n = 101), healthy controls (HC, n = 30) and Hodgkin’s Lymphoma (HL, n = 20) samples. (B) Proportion of samples that were positive for at least one autoantibody target in the CTD array (* = P < 0.05). A Fisher’s exact test was used to determine significance between groups. (C) Dot plots comparing MFI values for three of the most targeted autoantigens associated with myositis and overlap syndromes in iMCD patients. (D) Dot plots displaying three of the most commonly targeted autoantigens associated with SLE and SS in iMCD patients.

Figure 2

IgG ACAs are common in iMCD patients. (A) Heatmap displaying serum IgG ACAs identified by a 38-plex array of secreted and cell-surface proteins in iMCD patients (n = 101), healthy controls (HC, n = 30) and Hodgkin’s Lymphoma (HL, n = 20) samples. (B) Proportion of samples that were positive for at least one ACA. iMCD and Lymphoma cohorts were significantly enriched for ACAs (** = P < 0.01) compared to HC. A Fisher’s exact test was used to determine significance between groups. (C) Dotplots comparing MFI values for six cytokines and cell surface proteins that were most commonly targeted by ACA in iMCD patients. (D) Longitudinal analysis comparing flare and remission disease state MFI values. Only comparisons that surpassed a Bonferroni-corrected P-value cutoff after Wilcoxon signed rank tests are shown.