The Tumor Microenvironment of Nodular Lymphocyte Predominant Hodgkin Lymphoma Is a Unique Immunobiological Entity Distinct from Classical Hodgkin Lymphoma

There is proven pre-clinical and clinical efficacy of mono or combinatorial immune strategies to boost host anti-lymphoma immunity, with classical Hodgkin Lymphoma (cHL) seen as the 'poster child'. Approaches include blockade of immune-checkpoints on exhausted tumor-specific T-cells (via mAb blockade of PD-1, TIM3, LAG3, TIGIT or their ligands), activation of T-cells via mAbs agonistic to CD137, and finally modulation of FOXP3, CTLA-4 and/or LAG3 regulatory T-cells (Tregs) or immunosuppressive tumor-associated macrophages (TAMs). In contrast, studies characterizing the circulating and intra-tumoral microenvironment (TME) of the distinct but rare CD20+ Hodgkin Lymphoma entity (5-8% of HL), Nodular Lymphocyte Predominant Hodgkin Lymphoma (NLPHL), are minimal. Furthermore, to our knowledge no functional profiling studies comparing the host immunity of NLPHL with cHL has been performed.

We compared host immunity in 29 NLPHL patients, 30 cHL patients and 10 healthy individuals, with a focus on pertinent and clinically actionable immune parameters. Paraffin-embedded tissue and paired (pre- and post-therapy) peripheral blood mononuclear cells samples were interrogated by digital multiplex hybridization (Nanostring Cancer Immune Profiling Panel) and flow cytometry.

Although cytotoxic T-cell gene counts (CD8a, CD8b) were similar, compared to cHL there were higher levels of the immune effector activation marker CD137 (gene counts 439 vs. 287; P<0.01). Consistent with this, CD4 and the Treg markers LAG3, FOXP3 and CTLA-4 were lower in NLPHL (2-4 fold lower, all P<0.05), with no difference in T-helper cell activation markers CD40L and CD30L seen between tumors. TAMs and dendritic cell markers MARCO, CD36, CD68, CD163, COLEC12 and CD11b were all lower in NLPHL than cHL (all P<0.05). In line with the known 'rossette' formed around LP cells by PD-1+ T-lymphocytes, we observed strikingly elevated PD-1 and the other T-cell checkpoints TIM3 and TIGIT in NLPHL (all 2-3 fold, P<0.001). However, in line with the known gene amplification of PD-L1 on HRS cells and its presence on TAMs, gene counts of this checkpoint ligand were 2-fold higher in cHL (P<0.001).

Flow cytometry profiling of immune subsets in peripheral blood showed findings consistent with findings in the TME. Specifically, there was elevation of multiple exhaustion markers within CD4, CD8, and NK immune effector cells, with a striking proportion of highly anergic dual-LAG3/PD-1 positive CD8+ T-cells. Also there was elevation of immune-suppressive monocyte/macrophages in cHL relative to NLPHL.

Relative to healthy lymph nodes, there was prominent up-regulation of a range of T-cell associated exhaustion markers in both NLPHL and cHL, indicating dysregulated priming of effector immune responses and host immune homeostasis. Comparison between NLPHL and cHL illustrated that NLPHL had a myriad of features that marked its intratumoral TME as a unique immunobiological entity typified by elevated immune checkpoint markers and T-cells with a highly anergic phenotype.

Put together, these findings indicate that distinct immune evasion mechanisms are operative within the TME of NLPHL, including markedly higher levels of multiple immune-checkpoints relative to cHL. In contrast, Treg subsets and immune-suppressive monocyte/macrophages were relatively lower than that seen in cHL. T-cells frequently had dual immune-checkpoint expression. The findings from this study provides a compelling pre-clinical rationale for targeting PD-1 or combinatory checkpoint inhibition in NLPHL and sets the basis for future 'chemo-free' rituximab + checkpoint inhibitor clinical trials.