Expansion of Dysfunctional CD56-Negative NK Cells Is a Hallmark of NK Cell Activation in Ph⁺ Leukemia Patients Treated with Dasatinib

[Background] Tyrosine kinase inhibitors (TKIs), imatinib, nilotinib and dasatinib, are key drugs for the treatment of Philadelphia chromosome-positive (Ph⁺) leukemia. Dasatinib treatment markedly increases the number of large granular lymphocytes (LGLs) including NK cells in peripheral blood, which associates with a better prognosis. We previously showed that dasatinib expands CMV-associated highly differentiated NK cells in Ph⁺ leukemia patients through reactivation of CMV (Ishiyama, et al., Leukemia, 2016). NK cells consist of conventional CD56^bright NK cells and CD56^dim NK cells, and also rare CD56-negative (CD56^neg) NK cells, which have been reported to increase in chronic viral infection such as HIV and HCV. Here, we show that CD56^neg NK cells increase in CMV seropositive (CMV⁺) patients treated with dasatinib (DA), but not in those treated with other TKIs (OT). However, characteristics and clinical implications of CD56^neg NK cells in CMV⁺DA patients remain unknown. Therefore, we sought to examine the phenotypic and functional properties of CD56^neg NK cells in these patients.

[Methods] We assessed NK cell subsets in 36 DA and 26 OT patients. NK cells were defined as lin^-CD16⁺ or lin^-CD56⁺ cells in peripheral blood. NK-cell marker expression and cytotoxic activity were compared between CD56^neg NK cells and CD56^dim NK cells in CMV⁺ DA patients with 5% or more of CD56^neg NK cells. In vivo phosphorylation of intracellular signaling molecules in NK cells were evaluated by flow cytometry combined with phospho-specific antibodies (BD Phosflow^TM).

[Results] CD56^neg NK cells exclusively increased in CMV⁺ DA group, compared to CMV^- DA and CMV⁺ OT groups (median count: 20.9/μl, 1.1/μl, 1.5μl; p < 0.001. Median percentage: 4.5%, 0.8%, 0.4% of total NK cells; p < 0.001). CD56^neg NK cell counts strongly correlated with total NK cell counts in CMV⁺ DA (r = 0.720, p < 0.001). CD56^neg NK cells gradually increased over a period of one year after starting dasatinib. When we compared CD56^neg NK cells and CD56^dim NK cells in CMV⁺ DA, CD56^neg NK cells showed decreased expression of activating NK cell markers including NKG2C, CD57, NKG2D, NKp30 and NKp46. In contrast, expression of PD-1 increased. CD56^neg NK cells also showed lower rate of degranulation and IFN-γ production in functional assays. CMV⁺ DA patients with 4% or more of CD56^neg NK cells significantly achieved deep molecular responses at 1 year after starting dasatinib than those with lower CD56^neg NK cells (10/14, 3/14; p < 0.05). Phosflow analysis showed enhanced phosphorylation of ZAP70 in NK cells from CMV⁺ DA compared to CMV⁺ healthy controls (MFI/Isotype = 24.6 vs. 7.4, p < 0.01).

[Discussion] CD56^neg NK cells exclusively increased in CMV⁺ DA, and their increase persisted during dasatinib therapy. This suggests the similarity with the previous findings that CMV-associated differentiation in NK cells is enhanced during dasatinib therapy in CMV⁺ DA, which reflects the NK cell activation in response to CMV reactivation. In addition, CD56^neg NK cells exhibited downregulation of activating receptors, upregulation of PD-1, and lower cytotoxic activity, indicating that these cells are the anergic and exhausted population as seen in chronic viral infection. These findings suggest that CD56^neg NK cells accumulate owing to chronic stimulation by reactivated CMV during dasatinib therapy. ZAP70 is an immediate downstream tyrosine kinase of activating NK cell receptors, and is regulated by Src-family kinases, which is potentially inhibited by the kinase inhibitor property of dasatinib. Intriguingly, phosphorylation of ZAP70 in CD56^dim NK cells was enhanced in CMV⁺ DA, although they had taken dasatinib a few hours before collecting blood samples. This finding suggests that strong activation of NK cells by CMV reactivation in CMV⁺ DA may overcome the direct suppressive effect of dasatinib on NK cell activation.

[Conclusion] The accumulation of CD56^neg NK cells is a consequence of NK cell activation specifically observed in dasatinib-treated patients. The NK cell activation is likely to be a response to the CMV reactivation induced by dasatinib treatment. Assessing CD56^neg NK cells in peripheral blood could be a practical clinical indicator for the immunomodulatory effect of dasatinib, which significantly affects the prognosis.