Interferon-γ sparks inflammatory fire in Castleman disease Free

In this issue of Blood, Yin et al¹ shed further light on potentially pathogenic mechanisms in idiopathic multicentric Castleman disease (iMCD).

Castleman disease (CD) comprises a rare group of disorders with similar lymph node histopathology. Unicentric CD (UCD) is localized; multicentric CD (MCD) has more generalized lymphadenopathy. The multicentric form has 3 main varieties. Human herpesvirus 8–associated MCD is driven by viral interleukin-6 (IL-6) and occurs in the context of immunosuppression (eg, HIV). A second form co-occurs with POEMS (polyneuropathy, organomegaly, endocrinopathy, monoclonal plasma disorder, skin changes) syndrome and is called POEMS-associated MCD. Approximately 50% of patients with MCD have iMCD, the cause of which is unknown.

Clinically, iMCD is characterized by constitutional symptoms, abnormal labs (anemia, hypoalbuminemia, elevated erythrocyte sedimentation rate/C-reactive protein), and, in severe cases, cytokine-driven multiorgan failure and death. The iMCD spectrum is quite heterogeneous. A severe form of iMCD involves thrombocytopenia, anasarca, fever or reticulin fibrosis, renal failure, and organomegaly, and is called iMCD-TAFRO. Another form has a milder clinical phenotype, characterized by hypergammaglobulinemia, thrombocytosis, and plasmacytosis, and is referred to as idiopathic plasmacytic lymphadenopathy (iMCD-IPL). The remaining patients are classified as not otherwise specified (iMCD-NOS). Although the etiology of iMCD remains elusive, progress has been made in our understanding of the various immune cells involved and the intracellular signaling pathways activated. IL-6 has been linked to the pathogenesis of CD since as early as 1989 and was found to be produced by germinal center B cells. An antibody that directly neutralizes IL-6, siltuximab, the only treatment approved by the US Food and Drug Administration, showed efficacy in 34% of patients in a randomized controlled trial. Real-world data showed response rates varying from 33% to 81%.

In their study, Yin et al performed single-cell RNA-sequencing (scRNAseq) analysis of peripheral blood and lymph node samples of patients with iMCD. IL-6 was produced in lymph nodes by activated B cells, fibroblasts, and endothelial cells, whereas in peripheral blood it was mainly produced by activated B cells. Two recent studies also showed IL-6 expression in endothelial cells and in fibroblastic reticular cells.^2,3 The authors have described an important role for classical CCL monocytes, which expressed IL-6 receptors as well as large numbers of pro-inflammatory cytokines. Interferon-γ, produced by GZMK⁺ NK cells, NKT cells, and CD8⁺ Tex, contributed to the CCL-monocyte–induced cytokine storm, especially in patients with more severe iMCD. Overall, the work solidifies the important role of IL-6 in iMCD and points to the importance of monocytes and other inflammatory mediators such as interferon-γ. Mumau et al, using serum proteomics, also presented evidence of elevated interferon-γ activity in iMCD patients by measuring interferon-γ–inducible proteins such as IL-18–binding protein.⁴ IL-18–binding protein was increased during active iMCD and decreased after response to siltuximab-mediated IL-6 blockade. In their validation cohort, Yin et al show that high serum interferon-γ levels indicated more severe disease and worse outcomes. Together, these findings suggest that interferon-γ may serve as a biomarker of disease activity and can potentially be a new therapeutic target, especially in severely ill patients or anti–IL-6 monoclonal antibody (mAb) nonresponders. Yin et al also suggest a role for rituximab to eliminate IL-6–secreting B cells, especially in severely ill patients who may need therapy beyond anti–IL-6 mAb. Clinical evidence suggested rituximab is effective in a subgroup of iMCD patients. The findings of Yin et al could lead to the discovery of a biomarker that may help to identify those patients most likely to respond.

It is helpful to consider these findings in the context of other research insights related to iMCD pathogenesis across the various subtypes. Nishikori and colleagues recently compared the lymph node transcriptomes of iMCD-IPL and iMCD-TAFRO/NOS, focusing on cytokine storm–related genes.⁵ They found no major differences in IL-6 expression between the subgroups. However, the iMCD-TAFRO and iMCD-NOS patients had notably upregulated gene expression of IL18, IL18-BP, STAT3, CXCL13, PDGF, and VEGF as well as genes involved in MAPK, NF-κB, and mammalian target-of-rapamycin (mTOR) signaling compared with iMCD-IPL. The increased STAT3 gene expression was presumably via cytokines beyond IL-6 such as interferons. In this context, it is of interest to note that Pai et al⁶ also found, in patients with severe iMCD, increases in CD56 bright cytokine-secreting NK cells, pro-inflammatory monocytes, and activated CD8⁺ T cells expressing granzyme B and perforin, similar to findings of Yin et al. Furthermore, a type I interferon response gene signature was observed using scRNA and gene set enrichment analysis. Interferon-β–inducible mTOR signaling in CD8⁺ T cells and classical monocytes could be abrogated by both the mTOR inhibitor sirolimus and the JAK1/2 inhibitor ruxolitinib. mTOR activation has been found in iMCD lymph nodes and the mTOR inhibitor sirolimus has clinical activity in iMCD. Pierson et al discovered JAK-STAT activation in the lymph nodes of siltuximab nonresponders.⁷ A few iMCD patients responded to treatment with the JAK1/2 inhibitor ruxolitinib, and a clinical trial has been planned. In another recent study, Yin et al also drew attention to scRNAseq data showing that CXCL13 was overexpressed in Tph cells, which interacted with IL-6–activated B cells promoting IL-6 hyperresponsiveness.⁸ Harada et al found that Tph cells from iMCD-NOS patients were responsible for CXCL13 secretion in immunodeficient mice.⁹ Inhibition of CXCL13 abrogated the inflammatory response and improved mouse survival. Thus, interruption of the CXCL13/CXCR5 axis constitutes a potential therapeutic approach.

What are the clinical consequences of all these new insights? iMCD-IPL appears to have predominantly IL-6-signaling–mediated disease and responds extremely well to IL-6 mAb blockade. Patients with severe disease probably have activation of the JAK-STAT pathway through mechanisms not only confined to IL-6, which likely explains why some of these patients are susceptible to IL-6 blockade whereas others need cytotoxic chemotherapy or other mechanisms targeted. Multiple alternate pathways have emerged in these patients, providing new therapeutic targets such as interferon-γ, mTOR, JAK-STAT, tumor necrosis factor, and CXCL13/CXCR5. Recently, one highly refractory patient was found to respond to adalimumab, an anti–tumor necrosis factor mAb.¹⁰ The challenge will be to identify the efficacy of these potential agents and develop biomarkers to guide treatment in anti–IL-6 nonresponders.

Conflict-of-interest disclosure: F.v.R. serves as an advisory board member for Bristol Myers Squibb, the Castleman Disease Collaborative Network, GlaxoSmithKline, Janssen Pharmaceuticals, Kite Pharma, and Recordati; has received research funding for the ACCELERATE Registry; has received consulting fees from EUSA Pharma; and has received research study drug, with no associated research funding, for the clinical trial of sirolimus from Pfizer (NCT03933904). D.F. is an advisory board member for the Castleman Disease Collaborative Network; has received research funding for the ACCELERATE Registry; has received consulting fees from EUSA Pharma; has received research study drug, with no associated research funding, for the clinical trial of sirolimus from Pfizer (NCT03933904); and has 2 provisional patents pending related to the diagnosis and treatment of iMCD, including 1 related to CXCL13 as a biomarker in iMCD.