P-349: A retrospective assessment of biologically relevant genomic perturbations and variants of unknown significance in a cohort of plasma cell disorder at our institution; a single center experience

Introduction: Significant advances in understanding genomic profiles of plasma cell disorder patients have been made recently. However, the spectrum of genomic variants of uncertain significance (VUS) including their frequency and malignant transformation potential has not been well studied. In our cohort, we assessed the prevalence of biologically relevant variants and VUS using the Tempus xT® platform which is a targeted panel of 648 genes identified by DNA next-generation sequencing (NGS) enriched for clinically relevant and cancer driver genes and genes of emerging or uncertain clinical significance. We also assessed MSigDb (Human Molecular Signatures Database) hallmark gene set pathways that can be impacted by genomic VUS using a web-based tool Enrich R® providing various types of visualization summaries of collective functions of gene lists.

Methods: Genomic data from bone marrow aspirate samples of sixty-four patients in our institution with monoclonal gammopathies (MGUS), smoldering myeloma (SM), multiple myeloma (MM), AL amyloidosis (AMY) and Waldenstrom’s macroglobulinemia (WM) was abstracted using the Tempus xT® NGS platform and compared. Enrich-R® was then used to identify the top 3 Hallmark 2020 MSigDb pathways that were relevant amongst VUS.

Results: Amongst all patients, 55 biologically relevant mutations and 292 VUS mutations were identified. Twenty-seven of the biologically relevant mutations (49%) were in MM (24/64) with the median number per patient being 1 (0-4). The most frequently observed mutations were that of DNMT3A LOF (6, 22.22%) and KRAS GOF (3, 11.11%). 132 VUS mutations with median of 5 (0-23) mutations per patient were also identified of which KMT2D LOF (5, 3.79%) and NOTCH1 GOF (4, 3.03%) were the most frequently identified. Amongst patients with MGUS (17/64), 6 (12%) biologically relevant mutations were identified whilst 70 VUS mutations with median 4 (1-8) mutations per patient were also identified. Most frequent VUS were ARID1B LOF (4, 5.71%) and ATM LOF(2, 2.86%). Mean tumor mutational burden (TMB;m/mB) of all disease states was 1.67 in patients with AMY, WM (1.4), and MM (1.32) in comparison to that of MGUS (0.37) and SM (0.23) patients (p 0.08; Mann Whitney U). Identified VUS were input into curated hallmark MSigDb gene sets in EnrichR that identified the following targets in MM patients, E2F Target, G2-M Checkpoint and p53 pathway and in MGUS patents, PI3K/AKT/mTOR signaling, p53 pathway, and Estrogen response late in descending order of importance.

Conclusions: We herein report the genomic profile of plasma cell disorders using a commercially available assay at the time of diagnosis. Higher TMB and number of VUS per patient trended towards malignant disease states, suggesting that estimation of the known and unknown genomic burden can inform transformation potential. Pathway analysis of VUS genes can also be informative of their biological relevance if truly differentially expressed.