We recently demonstrated that ruxolitinib (INCB018424), the first approved JAK1/JAK2 inhibitor for treatment of myelofibrosis (MF), exerts potent anti-inflammatory activity. This may at least in part explain higher infection rates observed in ruxolitinib-treated patients. NK cells are critical for cancer-immune surveillance and cytokine-mediated signals are central for proper NK cell activation. We here aimed to characterize in detail the effects of JAK1/2 inhibition on human NK cells.


Highly purified CD56+ NK cells were isolated from human peripheral buffy coats by magnetic bead isolation and subsequently exposed to increasing concentrations of ruxolitinib (0.1-10 µM). Cytokine (1000U/ml IL-2, 25ng/ml IL-15)-induced NK cell proliferation was analyzed by CFSE dilution. Phenotypic and functional NK cell activation markers (NKp46, NKG2D, Granzyme B, CD16, and CD69) were analyzed by flow cytometry (including CD107a expression for degranulation). NK cell function was tested by flow-cytometry-based killing assays and quantification of IFN-γ production upon stimulation with either MHC class I-deficient K562 target cells or cytokines (IL-12, IL-18). In addition, phenotypic and functional analyses were also tested during NK receptor activation via plate-bound activating NKp46 antibodies. Signaling events were analyzed by Western Blot analysis to detect phosphorylation of JAK1 and JAK2 as well as by applying phospho-flow technology to evaluate ruxolitinib-mediated changes of cytokine-dependent signalling cascades (pS6, pSTAT1, pSTAT3, pSTAT5, pERK, pAKT, pP38, and pZAP70).


Our results demonstrate provide first evidence that ruxolitinib profoundly affects cytokine-induced NK cell activation. This includes a significant and dose-dependent reduction of NK cell proliferation, reduced induction of activation-associated surface markers (including NKp46, NKG2D, Granzyme B, CD16, CD69) as well as impaired killing activity against the classical NK target cell line K562. In addition, all main functional activities of NK cells are down-regulated as shown by reduced cytotoxic capacity, impaired degranulation and IFN-γ production. After wash-out, the inhibitory effects of ruxolitinib on NK cells are fully reversible, as shown by proper re-activation by cytokines. In contrast to cytokine-mediated NK cell activation, stimulation via the NK-specific receptor NKp46 are not affected by ruxolitinib. Of note, ruxolitinib does not affect NK cell viability. On a molecular level, phospho-flow analyses revealed that cytokine associated signaling events, such as phosphorylation of STAT5 and S6 were dose-dependently reduced by ruxolitinib in primary human NK cells.


Ruxolitinib strongly inhibits NK cell activation leading to impaired proliferation and functional activity. Experiments verifying these effects in patients are currently ongoing and will be presented at the meeting. Our findings may have important clinical implications, when considering the application of ruxolitinib as GvHD therapy, because NK cells are critically involved in the GvL effect after allogeneic stem cell transplantation.


Wolf:Novartis: Honoraria, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.