Co-Occurrence of MDS with MGUS Represents a Unique Disease Entity with Specific Molecular Phenotype and Potential Improved Clinical Outcome

Myelodysplastic syndrome (MDS) has been associated with co-occurring ailments of various origin. In particular, rheumatologic disorders (Felty's syndrome, VEXAS), malignancies such as LGL (large granular lymphocytosis) and pre-malignant conditions like MGUS (monoclonal gammopathy of undetermined significance) are among the most frequent. For instance, the co-occurrence of MDS and MGUS is identified in about 10-15% of all MDS presentations. MGUS is a disorder that almost always precedes the development of multiple myeloma and is characterized by the production of monoclonal M protein. The microenvironment created in the bone marrow of MDS patients may pave the way for the progression and development of inflammatory and neoplastic diseases (or vice versa) but the understanding of this pathogenesis is yet to be firmly elucidated.

We studied a cohort of 1627 MDS and secondary AML patients (median age 69 years, IQR:61-76) and retrospectively searched for the diagnosis of MGUS or multiple myeloma. The main aim of our study was to characterize the unique molecular, clinical, and phenotypic features of MDS patients who develop co-occurring plasma cell dyscrasias in an attempt to identify patients at higher risk for the development of this dyad as well as to investigate any underlying pathogenesis. Overall, we identified 97 (6%) patients with MGUS, 19 (1%) with multiple myeloma and 2 (0.1%) with smoldering myeloma. The median age of patients diagnosed with concurrent MDS/MGUS was 72.6 years and was significantly higher than that registered in the non-MGUS cohort (p=0.001). When expanding our investigation to patients with myeloid disorders of any type, patients diagnosed with concurrent myeloid neoplasms/MGUS (versus those without MGUS ) were found to more often have s-AML (31% vs 26%) and MDS-SLD (13% vs. 8%), and less likely to have MDS-MLD (12% vs. 18%), and had similar rates of MDS/MPN (15% vs.14%), MDS-5q (7% vs. 6%), RAEB-1 (7% vs. 7%), RAEB-2 (4% vs. 9%), CMML-1 (5% vs. 7%), and MDS-U (4% vs. 3%).

Complex karyotype was identified in 7% of patients diagnosed with MDS/MGUS (vs.18% in the non MGUS cohort, p=0.05). Deletion of the chromosome 20 was identified in 3% of MDS/MGUS cohort vs. 7% of non-MGUS cohort and 5% of MDS patients. Monosomies of either chromosomes 5 and 7 were not found to be different between the MDS and MGUS/MDS cohorts (p=0.412 and 0.963, respectively). Evaluation of the molecular profiles between patients diagnosed with concurrent MDS/MGUS versus MDS alone showed that patients in the MDS/MGUS cohort were more likely to have mutations of CEBPA (4% vs. 2%, p =0.05) and STAT3 (3% vs. 0.4% , p=0.001), and were not different as of mutation frequencies for the most commonly mutated genes such as DNMT3A (15% vs.11%), FLT3 (4% vs. 3%), IDH1 (6% vs. 3%), IDH2 (5% each), JAK2 (11% vs. 7%), RUNX1 (10% vs. 10%), SF3B1(14% vs. 10%), SRSF2 (16% vs. 13%), CUX1 (6% vs. 4%), DDX41 (3% vs. 0.5%) and TET2 (20% each).

When looking at disease outcomes, patients with MDS/MGUS were characterized by better survival as compared to those belonging to the non-MGUS cohort (median 13 vs 9 months, HR: 0.677, 95%CI: 0.525-0.872, p=0.003). In addition, a multivariate analysis model (including the variable mentioned) showed that the occurrence of the MGUS/MDS combination (HR: 0.710, 95%CI: 0.568-0.937, p=0.008), SF3B1 (HR: 0.7, 95%CI: 0.561-0.970, p=0.002), RUNX1 (HR: 1.5, 95%CI:1.236-1.945, p<0.001), TP53 mutation (HR: 1.4, 95%CI:1.196-1.750, p<0.001), and complex karyotype (HR: 2.3, p<0.001), were all significant predictors of overall survival.

In conclusion, based on our study MGUS/MDS overlap may be a unique disorder with noted molecular and phenotypic features. The overall survival of patients with MDS/MGUS appears to be better than that of non-MGUS/ MDS patients. Furthermore, the identification of patients with MDS/MGUS dyad may be under-estimated with a possible missed diagnosis of a co-existing condition and mis-classification of MDS patients. In particular, our results call for a broad screening for plasma cell dyscrasias in MDS populations, especially if the presence of such conditions might be a marker of improved outcome. Validation and etiology for this improved observed outcomes warrants further investigation in order to shed light onto this unique disease entity.