Atypical Stem Cell, Pre-B-Cell, T-Cell and Myeloid Gene Expression Characterizes Early Waldenstrom's Macroglobulinemia Clones Which Diminishes with Advancing Disease and Has Therapeutic Implications

Background: In previous studies using multi-omic data for 253 treatment-naive patients with Waldenstrom's Macroglobulinemia (WM), we identified three subtypes of WM: B-cell like (BCL), Plasma cell like (PCL), and an intermediate clone enriched for early/smoldering WM from which the other two evolved (Hunter et al, ASH 2022). The BCL and PCL subtypes show distinct mutation and transcriptional signatures, as well as clinical characteristics. Diffusion pseudo-time (DPT) analysis suggests all patient tumors, regardless of subtype have a shared evolutionary path and can be divided into Early and Late DPT determined stages that mirror disease progression. Additional analysis of DPT suggests it represents a continuous process akin to clonal evolution. We therefore sought to further characterize the DPT signal in MYD88 mutated WM.

Methods: RNASeq was performed on CD19-selected bone marrow (BM) samples from 249 of the 253 treatment-naive WM patients who were MYD88 mutated, as well as 13 paired CD19+CD27- and CD19+CD27+ selected healthy donor (HD) peripheral blood samples. Whole exome sequencing of CD19-selected BM cells along with CD19-depleted peripheral blood mononuclear controls was also performed for 215 of the WM samples. To assess BM histology, slides from 66 BM paraffin blocks from these WM patients were analyzed.

Results: WM patients had median age of 66 (range 31-95 years), BM involvement of 50% (5-95%) and serum IgM level of 3,162 (104-10,321 mg/dL). The median follow-up for the study cohort was 8.8 (range 0.2 - 33.8 years). Patients were stratified into 5 cohorts by DPT values. Analysis of BM involvement by DPT cohorts revealed a stepwise increase with the median involvement rising from 20% in the lowest to 80% in the highest cohort (p<0.0001). Increasing DPT predicted for shorter time from biopsy to first therapy (p=0.0003) but was not predictive of time from diagnosis to first therapy consistent with DPT alignment by stage of WM development. RNASeq analysis revealed high expression of genes atypical of mature B-cells that diminished with increasing DPT ( Figure 1). Among these were the myeloid chemo-attractants CXCL1, CXCL8 and CXCL12 which fell from a median of 4.4, 67.2 and 41.1 to 0.1, 3.6 and 0.8 transcripts per million (TpM) in the lowest and highest DPT cohorts, respectively (p<0.0001). Accordingly, all the top 20 gene ontology (GO) terms associated with DPT were related to activation, degranulation, and migration of myeloid cells and leukocytes. Pathological review of the 66 BM slides ranked by DPT, showed increased early myeloid cells and eosinophils clustering near small lymphoid aggregates, and as DPT increased, increased lymphocytes and mast cells with decreased myeloid cell involvement was observed. By RNASeq, pre-B-cell genes including MYB, the VDJ recombination genes RAG1, RAG2 and DNTT as well as the surrogate light chain genes IGLL1 and VPREB1 fell from 22.5, 10, 13.5, 61.1, 144 and 38.3 to 2.2, 1, 1.1, 9.1, 12.2 and 4.3 TpM, respectively (p<0.0001). T-cell and myeloid genes including CD4, CD8A, FCGR3A (CD16), CD14 and CD33 also decreased with increased DPT from 67.6, 7.1, 80.5, 200, and 23.2 to 2.4, 0.9, 1.7, 3.6 and 0.7 TpM, respectively (p<0.0005). A subclonal stem cell signature was also detected, including CD34 and KIT (CD117), which fell from 1.67 and 0.64 to 0.15 and 0.06 TpM, respectively (0<0.0001). CD34, KIT, preBCR, CD4, CD8A, FCGR3A, CEACAM8 (CD66b), and CD33 expression was confirmed by CD19+ light chain restricted flow cytometry in WM but not in HD samples. In addition, the MB and plasma blast genes PRDM1 and XBP1 were elevated in all samples regardless of DPT, consistent with a putative memory B-cell origin. Progression free survival (PFS) in response to proteasome inhibitors revealed strong stratification by subtype (p=0.001), as well as inferior PFS with BTK-inhibitors in low DPT patients (p= 0.044; Figure 2), suggesting less BTK dependence in the early DPT clones. Consistent with this finding, gene set enrichment identified increased BCR signaling with increased DPT.

Conclusions: This analysis suggests that stem cell program reactivation and atypical marker expression characterize early WM clones that are ultimately replaced by clones exhibiting more typical B-cell markers with disease progression. The findings may be relevant to treatment selection and provide a framework for investigating subtype and early/late evolutionary staging in the treatment approach to WM.