Bone Marrow Microenvironment Changes in Myelodysplastic Neoplasms and Its Relationship with Clonal Hematopoiesis and Disease Progression

Background: Immune dysregulation and somatic gene mutations are known prognostic factors in myelodysplastic neoplasms (MDS). During the evolution of the disease, a decrease in the immune surveillance is observed, with impaired cytotoxicity and a decrease in mature natural killer cells (NK), among other changes. Also, over expression of inflammatory cytokines, plays a role in the immune environment. The interaction between immune populations and the malignant clone is not entirely understood, especially in the bone marrow. The aims of this study were to characterize bone marrow (BM) immune populations, especially NK cells, and its correlation with the malignant clone.

Methods: We prospectively collected 70 BM samples. Fifty MDS without RS or 5q deletion, 11 cytopenia of unknown significance (CUS) and 9 healthy donors (HD). The aims were to analyze NK cells, defined as (CD45⁺CD3^-CD19^-) and differentiating an immature population (CD56^brigthCD16⁻NKG2A/CD94⁺KIR^+/−) from a mature one (CD56^dim CD16⁺NKG2A/CD94⁻KIR⁺). Their different receptors: activators (NKp46, NKp30, NKG2C, NKG2D, NKp44, DNAM) and inhibitors (TIGIT, NKG2A, Irp60, and PD1), as well as clonal cells expression of their ligands (HLA-ABC, MICA-B, CD155, PD-L1). Finally, KIR haplotype was analyzed, by NGS. Other populations included were myeloid-derived suppressor cells (MDSC), differentiating granulocytic (Gr-MDSC: CD11b⁺CD33⁺HLA-DR^-CD15⁺CD14^-) and monocytic (Mo-MDSC:CD11b⁺CD33⁺ HLA-DRlow/CD15⁺ CD14⁺). The analysis was performed, by flow cytometry, on BM fresh samples, following a membrane staining protocol with 10-colors/3 lasers. Molecular analysis was performed by NGS using the Oncomine Myeloid Research Assay included 40 genes. Finally, cytokines concentrations were stablished with the commercial kit ProcartaPlex^TM Multiplex immunoassay.

Results: A total of 70 subjects (median age of 77 years old and 45% female) were included (Figure 1). Although, total NK was similar among the three groups [MDS 0.17x10⁹/L ICUS 0.07x10⁹/L, HD 0.20x10⁹/L, p= 0.1204], we observed differences between NK subpopulations. An increase of mature NK cells CD56^dimCD16⁺ in CUS (76.16%) compared with MDS (11.58%) and HD (13.67%) [p= 0.0005] was observed. Regarding NK activating receptors, DNAM receptor was highly expressed on MDS (96.05%) and HD (90.98%), compared to CUS (67.79%) [p= 0.007], and the expression of KIR2DS4 was decreased in MDS (28.47%) compared to HD (79.4%) [p= 0.005], as well as NKp44 on MDS (0.19%) compared with HD (0.34%) [p= 0.03]. Interestingly, no expression of KIR2DS4 or NKp44 were observed on CUS. Regarding inhibitor receptors, Irp60 was increased on MDS (95.52%) compared to CUS (85.07%) and HD (80.2%) [p= 0.036] (Figure 2). Inhibitory KIR haplotype AA was observed on 13/47 MDS (27.56%). Finally, no differences were observed between Gr-MDSC [MDS 0.51%, ICUS 0.13%, HD 0.33%, p=0.35] or Mo-MDSC [MDS 1.03%, ICUS 0.25%, HD 0.7%, p=0.28] expression, among groups.

Eighty-four percent (42/50) of MDS presented mutations at diagnosis, with a median number of 3. Interestingly, 63.6% (7/11) of CUS presented mutations, 2 DMT3A (VAF < 5%), 1 TET2 (VAF 40%), 1 IDH1 and 1 SF3B1 (VAF 33%), among others. Within subjects presenting mutations, we observed differences on the expression of NKG2A, showing an increase on those with ≥3 mutations (37.23%) vs <3 mutations (35.3%) [p=0.02]. Moreover, differences were found on KIR2DL1 [≥3 mutations 6.66% vs <3 mutations 15.61%, p=0.05]. On the 7 clonal cytopenia of unknown significance (CCUS), no differences were found on the expression of NK receptors compared to idiopathic cytopenia of unknown significance (ICUS). Cytokine analysis on BM of 39 MDS, showed increased levels of IL-10 in HR compared to LR [2 pg/ml vs. 1 pg/ml, p=0.04], including 16 (53%) subjects with very high concentrations (> 40 pg/ml).

In addition, we analyzed NK cells of 14 MDS subjects who had a second BM sample, one year after diagnosis. We documented an increase in CD244 NK population [15.23% follow-up vs 3.61% diagnosis, p=0.05].

Conclusions: Our study suggests the presence of immunosuppressive profile of NK cells on MDS, with gain of expression of NKG2A and higher expression of IL10 on high risk MDS. Interestingly, 63.6% CUS subjects were CCUS and showed increase cytotoxic profile. The increase of CD244 expression, at one year control, may be related to an exhausted environment leading to immunosuppression and progression.

Molero Yordi:Oryzon Genomics: Consultancy. Salamero:Abbie, BMS, Novartis: Consultancy, Honoraria. Diez-Campelo:BluePrint: Membership on an entity's Board of Directors or advisory committees; Takeda: Honoraria, Membership on an entity's Board of Directors or advisory committees; Bristol Myers Squibb: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees.