Integrated multi-omic profiling reveals two biologically distinct subgroups of splenic marginal zone lymphoma with prognostic relevance
Bibliographic record
Abstract
Abstract Splenic marginal zone lymphoma (SMZL) is a rare B-cell malignancy mainly affecting the spleen, bone marrow, and peripheral blood. Clinical outcomes are variable, with potential transformation into aggressive large B-cell tumors with poor survival rates. Despite advances in targeted therapies, specific biomarkers are urgently needed to guide treatment as incidence continues to rise. This study aims to extend our knowledge of SMZL biology by integrating genetic, phenotypic, transcriptomic, and epigenetic data to establish more precise molecular classifications aimed at guiding personalized treatments. We defined epigenetically distinct subtypes of SMZL using DNA methylation array data of 142 patients divided into a discovery (n=86, 60%) and a validation cohort (n=56, 40%). K-means clustering of the top 2000 most variable CpG sites consistently identified two similar clusters in both cohorts. Bootstrapped univariate survival analyses revealed significant differences in time to first treatment (TTFT) between these clusters in both the discovery (p=.007) and validation cohorts (p=.024). Subsequent clustering on the entire cohort allowed us to classify the subgroups as SMZL-HR (high risk, n=58, 41%) and SMZL-LR (low risk, n=84, 59%), reflecting the observed differences in TTFT. Twelve clinico-biological features were significantly enriched in SMZL-HR cases, including female sex (p<.001) , IGHV1-2*04 usage (p<.001), gene mutations (KLF2 (p<.001), KMT2D (p=.0015), TRAF3 (p<.001), NOTCH2 (p=.015), BCL10 (p=.007)) and chromosomal alterations (del(7q), gain(3q), gain(12q) (all p<0.01)). SMZL-HR patients also had higher rates of therapeutic intervention (p<0.001), disease transformation (p=0.01), and mortality (p<0.001) compared to SMZL-LR patients. Tumour mutational burden (TMB) (p<0.001) and the fraction of the SBS40 (p<0.001) mutational signatures were also increased in SMZL-HR compared to SMZL-LR. In contrast, SMZL-LR was associated with MYD88 mutations (p=.02), Trisomy 12 and 3 (p<.001 and p=.02). We found that the DNA methylation-based proliferative history score epiCMIT was significantly higher in SMZL-HR than SMZL-LR patients (p<0.001). TMB was positively correlated with epiCMIT (r=0.35, p<.001), reinforcing the link between extensive tumor proliferative histories and the acquisition of somatic mutations. Telomere length (TL) data (median 3.1, range: 2.38-7.57 kb) showed a significant negative correlation with epiCMIT (R=-0.3, p=.001). Transcriptomic comparisons of SMZL-HR and SMZL-LR revealed 399 differentially expressed genes (232 under-expressed, 167 overexpressed; FDR<.05, log fold change >1.5). Gene set enrichment analysis highlighted pathways linked to elevated cell division, specifically E2F targets (NES=1.98, p<.01) and the G2M checkpoint (NES=2.07, p<.01), and KAMMINGA_EZH2 targets (NES=1.91, p=.006). Taken together, these results suggest that SMZL-HR clones have a history of and higher potential for cellular division, potentiated by EZH2 targets associated with chromatin modification/stabilization, providing enhanced cellular resilience against replicative stress. Univariate Cox regression analysis tested the impact 60 clinico-biological features on TTFT and overall survival (OS). SMZL-HR status (HR: 2.0, p=.0013) and epiCMIT >median (HR: 1.67, p=.014) were significantly linked to shorter TTFT (5 vs. 16 months). Additionally, 46% of patients were classified into a poor-risk “NNK-like” group and 20% into the “High-M” group, as defined by Bonfiglio and Arribas, respectively. Both groups were associated with shorter TTFT (HR: 1.57 and 1.9, p=0.002 and 0.008, respectively). Significant predictors of shorter OS included the SMZL-HR epitype (HR: 2.5, p=.025). SMZL-HR patients had significantly shorter TTFT regardless of their Bonfiglio/Arribas classification. Multivariate Cox analysis (including 130 patients with 91 events) with 4 covariates (SMZL-HR, epiCMIT, NNK-like, High-M), revealed that SMZL-HR was the only independent variable in the final model (HR: 2.63, p=.001). Overall, this study presents a comprehensive framework that integrates (epi)genomic data with survival analysis, identifying two distinct disease entities, each with a discrete biological and clinical landscapes. This enhanced understanding supports the potential for improved personalized treatment strategies as well as better prognostic assessment for patients with SMZL.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".