Prognostic Role of Adaptive Immune Microenvironment in Patients with High-Risk Myelodysplastic Syndromes Treated with 5-Azacytidine

Abstract

Background/objectives: There are limited data regarding immunohistochemical profiling of immune cells in bone marrow trephine biopsies of patients with high-risk myelodysplastic syndromes (HR-MDS).

Methods: We sought to objectively quantify, with the use of digital pathology, the density (cells/mm2) of the prominent adaptive immunity cell populations in sixty-four (64) bone marrow trephine biopsies of HR-MDS patients receiving 5-Azacytidine. We focused on CD3(+) T cells, CD8(+) cytotoxic T cells (Tc), helper T cells (Th), Foxp3(+) regulatory T cells (Tregs), CD20(+) B-cells and CD138(+) plasma cells and evaluated the presence and the number of lymphoid aggregates. A control group of twenty "non-MDS" patients was included in the study.

Results: We identified a significant decrease in adaptive immune cell densities in the HR-MDS patients compared to the non-MDS controls. Increased T and Th cell densities correlated with the response to 5-Azacytidine (5-AZA) treatment. Higher T, Tc, Th and plasma cells densities and low B, Tregs and Tregs/T cells ratios correlated with increased overall survival. Reduced Tregs, Tregs/T cells, Tregs/Tc and plasma cells showed improved leukemia-free survival. A modified IPSS-R (IPSS-R-I), combining the initial IPSS-R with the immune populations' parameters, improved overall survival and showed a double-fold increase in Cox calculated hazard ratios.

Conclusions: Immunohistochemical bone marrow immune profiling represents a powerful and easily useable tool for investigating the possible role of bone marrow immune microenvironment in the pathogenesis and progression of MDS, but also its association with the response to 5-AZA treatment and clinical outcomes.

Keywords: 5-Azacitidine; digital pathology; immunoenvironment; immunohistochemistry; myelodysplastic syndromes.

Figure 1

(A) The five areas (R1–R5) with the highest densities of CD8+ cytotoxic T cells were selected (2× magnification left, 20× magnification right). Our metric procedure included bone marrow cellularity normalization by deducting bone and fat tissue areas as well as any artifactual empty spaces. The density of each immune cell population was calculated as the mean of the five different areas normalized per mm² and 100% BM cellularity. (B) The same areas were chosen for the estimation of CD3, CD20, CD138 and FOXP3. These are some examples that illustrate the variability of the density (high and low) of these immune populations (40× magnification).

Figure 2

The densities (mean value/mm² of bone marrow) of the immune cell populations according to WHO-defined entities. Individuals from the control group are also included.

Figure 3

Kaplan–Meier curves showing the overall survival (upper panel) and leukemia-free survival (lower panel) according to both the IPSS-R groups and the newly proposed IPSS-R-I groups.