Unveiling Transformation Mechanisms of Nodal Marginal Zone Lymphoma Using a Multi-Omics Approach

Introduction: Approximately 8% of nodal marginal zone lymphoma (NMZL) patients experience transformation into diffuse large B-cell lymphoma (DLBCL) within 5 years from diagnosis (PMID 37658062). Transformation of NMZL (tNMZL) is associated with a significantly worse prognosis, but the underlying mechanisms are unknown. In this study, we conducted a multi-omics analysis to explore the differences in gene expression and mutation profile between NMZL, tNMZL, and nodal DLBCL.

Methods: We analyzed 32 paraffin embedded tissue samples from 26 (t)NMZL patients, including 17 samples collected at initial NMZL diagnosis (7 of which transformed) and 15 samples collected post-transformation. For 6 patients, tissue samples were available both before and after transformation. As a comparator, 15 nodal DLBCL cases matched for cell-off-origin (COO), age, and sex were analyzed. Targeted sequencing was performed using a custom-made panel including 128 B-cell lymphoma-relevant genes (BLYMFv2 panel, an updated version of the LYMFv1 and BLYMF200 panels, PMID: 34478526, 34933651). Gene-expression profiling (GEP) was performed using the BLYM-777 panel (PMID: 35454765) including 777 genes covering 13 GEP signatures and capturing many aspects of B-cell lymphomagenesis (e.g. tumor microenvironment (TME), immune surveillance, and MYC signature).

Results: Median age at NMZL, tNMZL and nodal DLBCL diagnosis was 64 (48 – 82), 66 (53 – 82), and 64 years (46 – 75), respectively. In tNMZL, median time to transformation was 10.4 months (0 – 62.2). In total, 33 samples showed a sufficient sequencing quality (13 NMZL, 10 tNMZL, and 10 nodal DLBCL). The median number of pathogenic mutations was slightly higher in tNMZL (6.5 (2 – 16) in a total of 41 genes) compared to NMZL (3 (0 – 10) in a total of 33 genes), although this difference was not statistically significant (p = 0.10). Notably, 48% of the genes mutated in the NMZL samples were also mutated in the tNMZL samples. In 4 paired (t)NMZL cases, the median number of mutations at NMZL diagnosis was 2 (2 – 3) compared to 7.5 (2 – 16) at transformation, with an increase in number of mutations in 3 cases. In nodal DLBCL, the number of mutations (8.5 (5 – 17)) was higher compared to both tNMZL (p = 0.04) and NMZL (p < 0.01). Genes that were recurrently mutated at tNMZL diagnosis but not at NMZL diagnosis included B2M (30%), PIM1 (20%), CCND3 (20%), NOTCH2 (20%), POU2F2 (20%), IRF4 (20%), and BCL6 (20%). Of these genes, B2M (15%), PIM1 (15%), and NOTCH2 (15%) were also mutated in at least 2 nodal DLBCL cases.

In total, 46 samples were of sufficient quality for GEP (16 NMZL, 15 tNMZL, and 15 nodal DLBCL). GEP did not reveal any significant differences in gene expression levels, and also not in any of the 13 gene signatures between NMZL samples and tNMZL samples, nor between the paired NMZL and tNMZL samples. A significantly higher expression was observed for 113 genes, including MAGEA6, CCL20, CXCL8, and IL2, when comparing tNMZL to nodal DLBCL (log2 fold change > 1.5, P.adj < 0.05). In contrast, only three genes, i.e. TCIRG1, LIMD1, and COL1A2, had a significantly higher expression in tNMZL (log2 fold change < -1.5, P.adj < 0.05). Furthermore, the “Depleted” TME signature (based on Kotlov lymphoma TME signatures) was observed more frequently in tNMZL cases (10 out of 15 (66.7%)) compared to nodal DLBCL cases (2 out of 15 (13.3%)). The remaining 5 tNMZL cases were defined as “Mesenchymal”.

Conclusion: Our study provides insight into the molecular alterations associated with transformation of NMZL. Aside from a single mutation in TP53, we did not detect mutations in known driver genes responsible for transformation of indolent lymphoma in the diagnostic samples. However, we did identify several genes specifically mutated in tNMZL, including B2M, PIM1, and NOTCH2, that were also detected in nodal de novo DLBCL, suggesting their involvement in later stages of transformation. An increase in the number of mutations in 3 of 4 paired samples suggests a potential accumulation of genetic alterations over time. Gene-expression profiles did not show significant differences between NMZL and tNMZL, indicating a stable expression of B-cell and TME profiles.