Low Expression of IL-15 in Adult B-Lineage Acute Lymphoblastic Leukemia Is Associated with Central Nervous System Involvement at Initial Presentation.

Abstract

The biologic mechanisms of the development of central nervous system (CNS) involvement in acute lymphoblastic leukemia (ALL) is unknown. Risk factors associated with CNS involvement include the immunophenotype (e.g. T-ALL), high white blood cell (WBC) count and cytokine expression. Among the latter, interleukin (IL)-15 has shown to enhance the proliferation of both normal and malignant lymphocytes, thus, suggesting its potential role in leukemogenesis. It has been shown that in childhood ALL, CNS involvement is associated with higher IL-15 expression (Cario et al. 2006, Blood 108a:2270). In this study, we analyzed the expression of IL-15 and its alternatively spliced variants in leukemic cells from adult ALL patients with CNS involvement (CNS+) at initial presentation (n=31) and at first relapse (n=8), as well as in patients without CNS involvement at diagnosis (CNS−; n=57). Experiments were performed by real-time RT-PCR and the values were presented ratios comparing IL-15 to expression of the reference gene belta-actin. mRNA expression levels were also correlated with protein expression by western blot analysis. We found that the expression of total IL-15 was significantly lower in ALL patients with CNS+ at initial presentation (n=31, median=0.02, range 0–0.45, P<0.001) and at first relapse (n=8; median=0.03, range 0–0.08, P=0.019), in comparison to ALL patients without CNS (n=57, median=0.08, range 0–0.91). Similarly, lower expression values of IL-15 were found in B-cell precursor (BCP)-ALL (P=0.015) as well as in the subtype of common ALL (p=0.013) with CNS+, as compared to cases without CNS-. In CNS+ ALL, patients with BCR–ABL+–BCP–ALL (n=9) had lower IL-15 expression compared to patients with BCR–ABL–BCP–ALL (n=21, P=0.017). In contrast to CNS– patients, no statistically significant difference was found regarding IL-15 expression between BCR–ABL+- and BCR–ABL–BCP–ALL. Furthermore, the expression of IL-15 was more than 5-fold higher in T–lineage ALL (n=23) than in BCP-ALL (n=72, p<0.001). Among T-lineage-ALLs, CD1+ cortical-T-ALL strongly expressed IL-15 as compared with pre-T or mature T-ALLs. The expression of both spliced variants of long signal peptide (LSP)-IL-15 and short signal peptide (SSP)-IL-15 did not differ between CNS+ and CNS– of all cases. Interestingly, the expression of LSP-IL-15 and SSP-IL-15 is higher in common or pre-B than in pro-B-leukemic cells, whereas high LSP-IL-15 was found in cortical T-ALL, but not in pre-T or mature T-ALL. In conclusion, lower expression of IL-15 in adult BCP-ALL at diagnosis was associated with CNS involvement. This unexpected difference in IL-15 expression between adult and childhood ALL may reflect differences in biologic features of leukemic cells and/or inflammation processes in the pathogenesis of CNS disease. Furthermore, the expression of IL-15 and its spilced variants was correlated with lineage commitment and differentiation status of leukemic cells in B-lineage-ALL as well as in T-ALL. It remains to be evaluated whether these prognostic and biologic findings of distinct expression pattern of IL-15 in adult ALL subtypes will have therapeutical implications for the future antileukemic strategies.