Exploring the Cell-Free Transcriptome As Biomarker Source in IgM Gammopathies: New Insights from the Fondazione Italiana Linfomi “BIO-WM” Trial

Introduction

IgM gammopathies encompass a heterogeneous group of hematological conditions, ranging from IgM monoclonal gammopathies of uncertain significance (IgM-MGUS) to asymptomatic Waldenström’s Macroglobulinemia (aWM) and symptomatic WM (WM). This study aimed to investigate the potential of cell-free RNA (cfRNA) as a biomarker source in IgM gammopathies.

Methods

Diagnostic blood plasma samples were collected from a retrospective and prospective series of IgM gammopathy patients enrolled in the Fondazione Italiana Linfomi (FIL) “BIO-WM” trial (NCT03521516). The blood plasma was analyzed of 60 patients with IgM-MGUS (n=15), aWM (n=32), and WM (n=13), along with healthy controls (HC) (n=28), collected using EDTA or Cell-Free DNA BCT (Streck) tubes. RNA was extracted from 200 µL of plasma using the miRNeasy serum/plasma kit and sequenced on a NovaSeq 6000 instrument with the SMARTer Stranded Total RNA-seq pico v3 library preparation kit. Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathways enrichment analyses were performed using the clusterProfiler and ReactomePA packages, respectively. Bulk deconvolution was performed using the Tabula Sapiens v1 basis matrix. Flow cytometry and MYD88^L265P mutational dPCR data were also available.

Results

Due to superior quality of cfRNA in EDTA tubes (median number of genes with a count >10: n=4338 for EDTA versus n=35 for Streck) only data from 49/60 patients (12 IgM-MGUS, 30 aWM, 7 WM) and 14 HC were reported. The Streck tube’s poor performance was observed in all the samples of both patients and HC, so it was not influenced by other preanalytics before cfRNA extraction.

Compared to IgM-MGUS and HC, higher cfRNA concentrations were found both in aWM (p=0.037, 95% CI [3.5, 250 pg/mL]; p=0.039; 95% CI [2.3, 240 pg/mL]) and WM patients (p=0.036, 95% CI [30, 470 pg/mL]; p=0.031, 95% CI [13, 450 pg/mL]). CfRNA concentrations significantly correlated with IgM serum levels (p=0.0047, R=0.40) and bone marrow infiltration (p=0.0019, R=0.43).

Numerous differentially abundant coding and non-coding genes (DAGs) were identified in the subgroups. A comparison between aWM and WM to IgM-MGUS patients revealed 259 and 510 DAGs, respectively, including RASSF6 and G0S2, which have been previously reported in WM. DAG are often lymphoma related genes and seem to capture disease related biological differences between the groups. Of note, aWM and WM patients showed enrichment in neurodegeneration, platelet activation, signaling, and aggregation, as well as neutrophil degranulation pathways. Interestingly, MYD88^L265P plasma levels showed a weak negative correlation with normalized cfRNA counts of MYD88 (p=0.015, R=-0.36). TREML1, ITGA2B and PF4V1 are cfRNA transcripts that show high potential for classifying between HC, IgM-MGUS and (a)WM. However, more stringent analysis identified PF4V1 as a potential single marker to distinguish between HC, IgM-MGUS, and aWM/WM, with One-vs-Rest multiclass ROC AUCs of 0.98, 0.82, and 0.88, respectively.

Survival analysis of aWM/WM patients showed a higher blood plasma abundance of a 6-gene unfavorable signature (ECH1, EIF3B, DHX9, EPRS1, ATP5PF, HIST1H1C) linked to all-cause mortality (p<0.0001). Finally, a 3-gene unfavorable signature (PAFAH1B1, ARAF, SMG7) was associated with time to relapse or progression (p<0.0001).

Conclusion

This is the first reported evidence that a cell-free transcriptome signature observed in EDTA blood plasma can differentiate between distinct subgroups of IgM gammopathies. Moreover, cfRNA might assist in prognostication and therapy response prediction and elucidate underlying biological pathways. Further studies are warranted to confirm its role as a less invasive alternative to bone marrow examination.