Large granular lymphocyte (LGL) leukemia is a semiautonomous proliferation of cytotoxic T cells (CTL) often accompanied by lineage-restricted immune cytopenias, including neutropenia and anemia. Clonal expansion of CD57+ CTL in LGL leukemia may constitute an extreme pole of a spectrum of immune-mediated responses that range from polyclonal such as in aplastic anemia, likely with multiple clones and target antigens, oligoclonal as in some immune-mediated cytopenias, to clonally-restricted LGL lymphoproliferations. Consequently, LGL leukemia may be a clonally exaggerated response to chronic antigenic stimulation. NKG2D is present on reactive CTL, the normal counterparts of leukemic LGLs. Its ligand, MICA, is a tightly regulated, non-peptide-presenting HLA-related molecule induced during viral infections or cellular stress. As a mechanism of escape from anti-tumor cellular effector mechanisms, soluble MICA (sMICA) is secreted by various epithelial tumors. Similarly, herpes and adenoviruses have developed mechanisms for cellular MICA sequestration or secretion as means to evade immune-mediated elimination of infected cells. We thus hypothesized that NKG2D-MICA signaling plays a role in cytopenias associated with LGL leukemia.

We studied a cohort of 47 well annotated patients with LGL leukemia. Flow cytometric analysis showed that while CD57 expression on CD8 cells was expectedly higher in LGL than controls (p<0.001), there was no difference in the level of expression of NKG2D within the leukemic and normal CD57+ CTLs (p=0.576). When we examined MICA expression using flow cytometry, granulocytes derived from LGL patients showed increased signal intensity of MICA on CD15 cells from patients with LGL and neutropenia than in control granulocytes (mean fluorescent intensity 3.88 vs. 2.26; p=0.033). No significant difference was found when all LGL patients, including those in remission or patients with anemia as a hematologic presentation were investigated (p=0.102). Moreover, absolute neutrophil count was inversely correlated with MICA expression (R=0.50, p=0.035). In agreement with flow cytometry results of MICA levels, sMICA levels were found significantly elevated in LGL patients in comparison to controls in whom MICA was virtually absent (p<0.001). Overall elevated sMICA levels were present in 80% of patients. Based on the hypothesis that allelic variants of MICA may influence cytotoxicity mediated by NKG2D-expressing CTL we studied prevalence of MICA types among patients. Sequencing of germ line DNA from 28 patients revealed a higher frequency of MICA00801/A5.1 compared to matched controls (p=0.020) with a frequency of 64% in patients and 25% in controls. Moreover, 39% of patients were homozygous for this allele compared to 8% of controls. This allele also contains the single nucleotide polymorphism (SNP) rs1063635, previously linked to disease risk by 250K SNP-A analysis with predictive value in both the training set (PPV=56%, NPV=89%) and test set (PPV=64%, NPV=86%). The frequency of this SNP in homozygous form was 12% vs. 60% in LGL (p=.010, OR=9.1). However, in other control populations the frequencies of this polymorphism and MICA 00801/A5.1 was reported to be as high as 37.7% in the general population necessitating further studies to confirm/disprove this linkage. Nevertheless, our results imply that MICA-NKG2D signaling plays an important role in the etiology of LGL leukemia either by active expression of MICA as a result of autoimmune attack on neutrophils and granulocyte precursors, or that MICA induction constitutes a marker of a pathologic process rendering myeloid cells susceptible as targets for clonal LGLs. Active secretion of this same marker may reflect an underlying causative process evading complete immune clearance through mimicry of the very mechanism by which clinical cytopenias are manifested.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author