Nodular Lymphocyte Predominant Hodgkin Lymphoma with Splenic Involvement Is Characterized By Inflamed Tumor Microenvironment, High Expression of Checkpoint Molecule Gene-Signature and Adverse Outcome

Background: Nodular lymphocyte-predominant Hodgkin lymphoma (NLPHL) is a rare B-cell malignancy characterized by infrequent neoplastic cells embedded in an immunologically active tumor microenvironment (TME). Histologically, NLPHL can be divided into classical (Fan A-B) and variant (Fan C-F) growth patterns. Clinical course is generally indolent, and the majority of patients present with limited stage disease. However, advanced stage with splenic involvement and variant histology are known to be associated with more aggressive disease. The molecular features underlying this association remain unknown. To gain further insights into disease biology, we recruited NLPHL cases as part of the Atlas of Blood Cancer Genomes (ABCG) initiative, a consortium consisting of 26 institutions.

Design: We collected comprehensive clinicopathological data including splenic involvement from 135 NLPHL patients, with centralized review performed by a panel of dedicated hematopathologists to ensure accurate diagnosis per 2016 WHO criteria. We performed RNA sequencing on formalin-fixed paraffin-embedded (FFPE) diagnostic tumor lymph nodes (n=100) and enumerated tumor-infiltrating immune cell compositions using FARDEEP with signature matrix LM22 from CIBERSORT. In addition, we performed gene set enrichment analysis to compare molecular pathways between the patients with and without splenic involvement. The Kaplan-Meier method was used to calculate overall survival (OS) and progression free survival (PFS) probabilities, and the log-rank test was used for comparisons between groups. Continuous variables were compared using the Wilcoxon rank-sum test.

Results: The median age at diagnosis was 42 years (range 5-80 years). Most patients were males (n=96; 71%) and had limited stage disease (n=96, 74%). Splenic involvement was present in 17 patients (13%). Histologic growth pattern was available for 97 cases, 41 (42%) with variant and 56 (58%) with classical histology. Variant histology was enriched in the patients with splenic involvement. Thirty-five (28%) relapses and 5 (3.9%) transformations occurred during a median follow-up of 80 months (range 0.36-604 months). Ten-year OS and PFS rates for the whole cohort were 86% (95% CI, 78-95%) and 59% (95% CI, 50-71%), respectively. Patients with splenic involvement at diagnosis tended to have inferior outcome at 10 years compared with those without splenic involvement (PFS 40% vs 62%, P=0.129; OS 79% vs 88%, P=0.0553).

According to in silico immunophenotyping, NLPHLs with splenic involvement were characterized by a lower proportion of naïve B cells and increased proportions of CD8+ T cells and M1-macrophages compared to the NLPHLs without splenic involvement (P<0.001 for all comparisons). Moreover, splenic involvement was associated with higher expression of checkpoint protein-encoding genes, including PDCD1 (PD-1), CD274 (PD-L1), CTLA4, LAG3, HAVCR2 (TIM-3) and TIGIT (Figure 1; P<0.001). Finally, we found a strong association between splenic involvement and gene expression profile related to inflammatory response (P<0.001), response to interferon gamma (P<0.001) and complement pathways (P<0.001).

Conclusion: Our study represents the largest comprehensive clinical and transcriptomic analyses of NLPHL to date. The results demonstrate that NLPHLs with splenic involvement are characterized by inflamed TME with high expression of checkpoint molecule gene-signature and are associated with inferior outcomes.

Figure 1. Unsupervised clustering for expression of checkpoint genes stratified by spleen involvement. Samples are ordered by the sum of normalized checkpoint gene expression values (checkpoint score)