Abstract
Malignancy-associated secondary hemophagocytic lymphohistiocytosis (sHLH), particularly in the context of lymphoma, is a life-threatening hyperinflammatory syndrome with a dismal prognosis. While monocytes are recognized as central drivers of HLH, their transcriptional programs and developmental origins in lymphoma-associated sHLH remain largely unknown. We hypothesized that aberrant myelopoiesis in this setting gives rise to dysfunctional monocytes that fuel both immune suppression and sustained inflammation.
To test this, we performed single-cell RNA sequencing (scRNA-seq) on paired bone marrow (BM) and peripheral blood (PB) mononuclear cells from patients with B-cell lymphoma either with sHLH (n=14) or without sHLH (non-HLH, n=7). To distinguish sHLH-specific changes from those associated with inherited cytotoxic defects, we included primary HLH patient samples carrying mutations in PRF1 (n=2) or UNC13D (n=4) as disease controls. In total, 110,943 cells were transcriptionally profiled.
Cell composition analysis revealed marked expansion of monocytes in both BM and PB of sHLH patients. These monocytes exhibited profound transcriptional alterations and segregated into two principal differentiation trajectories, both originating from an immature, bipotent monocyte population arrested at a critical developmental checkpoint. This bipotent population aligned transcriptionally with the monocyte state 1 (MS1) program, characterized by elevated expression of S100A8/9, IL1R2, and PLAC8, coupled with low expression of HLA-DR, defining an immune-suppressive phenotype within the hyperinflammatory milieu of lymphoma-associated sHLH.
This reprogramming was rooted in the BM. We observed robust signatures of emergency myelopoiesis in sHLH, including expansion of myeloid-biased hematopoietic stem cells (HSCs) and granulocyte-monocyte progenitors (GMPs). These sHLH-derived progenitors were enriched for inflammatory response pathways, particularly TNF-α and IFN-γ. Notably, the dysfunctional MS1 signature was already imprinted at the GMP stage, indicating pre-programmed monocytic dysregulation.
Among the two terminal monocyte fates, one trajectory (lineage 2) was preferentially expanded in sHLH. This lineage was characterized by a complex immunomodulatory profile, simultaneously expressing pro-inflammatory mediators (CXCL2, CXCL3, IL6), the anti-inflammatory cytokine IL10, and myeloid-derived immune checkpoint ligands implicated in T cell suppression (VSIG4, LILRB4). Communication network analysis predicted that this lineage 2 monocytes reinforce the MS1 program in bipotent monocytes through IL-6 and IL-10 signaling, supported by selective activation of the STAT3 regulon in MS1-like monocytes.
This myeloid dysregulation was associated with widespread immune paralysis. The abundance of MS1-like monocytes correlated inversely with CD4⁺ and CD8⁺ T cell frequencies in sHLH patients. In parallel, we observed a reduced frequency of conventional dendritic cells, impaired MHC class II antigen presentation, and enrichment of inflammatory signaling pathways. Concurrently, CD4⁺ and CD8⁺ T cells in sHLH exhibited metabolic hyperactivation, stress response upregulations, and expression of multiple inhibitory receptors, including HAVCR2, PDCD1, and LAG3. However, the absence of TOX upregulation suggests a state of dysfunctional activation rather than classical exhaustion.
Our findings delineate an immunopathogenic cascade in lymphoma-associated sHLH. Emergency myelopoiesis skews GMP differentiation to generate dysfunctional MS1-like monocytes. In parallel, an immunomodulatory monocyte population (lineage 2) sustains these MS1-like cells through IL-6/IL-10/STAT3 signaling, forming a self-reinforcing loop that amplifies innate inflammation and impairs adaptive immunity. Our study provides a mechanistic framework for immune dysregulation in sHLH and highlights emergency myelopoiesis and monocytic STAT3 signaling as potential therapeutic targets.
