Dissection of the Effects of JAK and BTK Inhibitors on the Functionality of Healthy and Malignant Lymphocytes

Tom Hofland, Iris de Weerdt, Hanneke ter Burg, Renate de Boer, Stacey Tannheimer, Sanne H. Tonino, Arnon P. Kater and Eric Eldering

J Immunol published online 11 September 2019

 Subscription Permissions Email Alerts
Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription
Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

Published September 11, 2019, doi:10.4049/jimmunol.1900321

The Journal of Immunology

 

Dissection of the Effects of JAK and BTK Inhibitors on the Functionality of Healthy and Malignant Lymphocytes

Tom Hofland,*,†,1 Iris de Weerdt,*,†,1 Hanneke ter Burg,*,† Renate de Boer,*,† Stacey Tannheimer,‡ Sanne H. Tonino,†,x Arnon P. Kater,†,x and Eric Eldering*,x
Despite the emergence of small molecule inhibitors, current treatment strategies for chronic lymphocytic leukemia (CLL) are not curative, and the search for new therapeutic modalities continues. Prosurvival signaling derived from the microenvironment is often mediated via JAK signaling. However, whether JAK inhibitors are useful in CLL therapy has not been studied extensively. JAK inhibitors are valuable therapeutic agents in myelofibrosis and show promising results in graft-versus-host-disease. However, JAK inhibition is associated with an increased infection risk, presumably because of the effect on other immune cells, a feature shared with other kinase inhibitors used for CLL treatment, such as the BTK inhibitor ibrutinib and the PI3Kd inhibitor idelalisib. We compared functional effects of the JAK1/2 inhibitors momelotinib and ruxolitinib, the BTK inhibitors ibrutinib and tirabrutinib, and PI3Kd inhibitor idelalisib on malignant CLL cells but also on healthy human T, B, and NK lymphocytes. We found several interesting differences among the inhibitors, apart from expected and well-known effects. Momelotinib but not ruxolitinib blocked cytokine-induced proliferation of CLL cells. Momelotinib also reduced BCR signaling, in contrast to ruxolitinib, indicating that these JAK inhibitors in fact have a distinct target spectrum. In contrast to tirabrutinib, ibrutinib had inhibitory effects on T cell activation, probably because of ITK inhibition. Remarkably, both BTK inhibitors stimulated IFN-g production in a mixed lymphocyte reaction. Collectively, our results demonstrate that kinase inhibitors directed at identical targets may have differential effects on lymphocyte function. Their unique profile could be strategically employed to balance desired versus unwanted lymphocyte inhibition. The Journal of Immunology, 2019, 203: 000–000.
hronic lymphocytic leukemia (CLL) is characterized by the accumulation of malignant B cells in blood, lymph nodes, and bone marrow (1). Despite the development of
targeted compounds and immunotherapies to eradicate CLL cells, none of the current treatments for CLL is curative. The Bcl-2 inhibitor venetoclax and Bruton tyrosine kinase (BTK) inhibitor ibrutinib have shown substantial clinical efficacy in most CLL patients (2, 3). However, mutations leading to drug resistance have already been described for both drugs, and, because of the high economic burden of lifelong treatments, alternative strategies leading to actual curative treatments should still be pursued (4–7). Most treatments in CLL are counteracted by the tumor-supportive microenvironment in secondary lymphoid organs, where CLL cells

*Department of Experimental Immunology, Amsterdam Infection and Immunity Institute, Amsterdam University Medical Centers, University of Amsterdam, 1105 AZ Amsterdam, the Netherlands; †Department of Hematology, Cancer Center Amsterdam, Amsterdam University Medical Centers, University of Amsterdam, 1105 AZ Amsterdam, the Netherlands; ‡Gilead Sciences, Foster City, CA 94404; and xLymphoma and Myeloma Center 1105 AZ Amsterdam, Amsterdam, the Netherlands
1T.H. and I.d.W. contributed equally.
Received for publication March 18, 2019. Accepted for publication August 10, 2019.
Address correspondence and reprint requests to Tom Hofland or Prof. Eric Eldering, Department of Experimental Immunology, Hematology, Amsterdam University Med- ical Centers, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands (T.H.) or De- partment of Experimental Immunology, Lymphoma and Myeloma Center, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands (E.E.). E-mail addresses: [email protected] (T.H.) or [email protected] (E.E.)
The online version of this article contains supplemental material.
Abbreviations used in this article: Amsterdam UMC, Amsterdam University Medical Centers; BTK, Bruton tyrosine kinase; CLL, chronic lymphocytic leukemia; GvHD, graft-versus-host disease; ITK, inducible tyrosine kinase; moDC, monocyte-derived dendritic cell; PI, propidium iodide.

Copyright © 2019 by The American Association of Immunologists, Inc. 0022-1767/19/$37.50 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1900321

receive prosurvival signaling from surrounding stromal cells, T cells, and macrophages (1, 8–10). Therefore, disrupting prosurvival sig- naling within the microenvironment is a sensible therapeutic aim in CLL treatment. T cell–derived IL-4 and IL-21 are known survival factors that both signal via JAK (11, 12). Inhibitors of JAKs have been developed, most notably momelotinib and ruxolitinib (both targeting JAK1 and JAK2). Ruxolitinib has been approved by the Food and Drug Administration and European Medicines Agency for the treatment of myelofibrosis. Therapeutic inhibition of JAKs with these inhibitors has shown significant clinical re- sponse in myelofibrosis patients, leading to reduced spleen sizes and improved overall survival (13–16). More recently, JAK in- hibitors have also shown promising clinical results in graft- versus-host disease (GvHD), leading to a reduction of steroid need (17, 18). In CLL, the biological activity of JAK inhibitors is demonstrated by lymphocyte redistribution out of the lymph nodes and disease stabilization. Although JAK inhibitors lack efficacy as monotherapeutic agents, based on the prosurvival contribution of IL-4 and IL-21, combination strategies involving JAK inhibitors may improve clinical responses in CLL treatment (19–21).
As JAK signaling plays a central role in the function of many immune cells, it is not surprising that JAK inhibitors have been shown to modulate the function of T cells, NK cells, and dendritic cells (22–25). In addition, other kinase inhibitors currently used or studied for CLL treatment also display side effects by the inhibition of off-target kinases. For example, ibrutinib has well- documented off-target effects on T cells by binding to IL-2– inducible tyrosine kinase (ITK) (26, 27). Idelalisib, a PI3Kd inhibitor used to treat refractory CLL patients, alters T cell function through modulation of TCR signaling, possibly explaining the in- creased amount of atypical infections and autoimmune complica- tions observed in patients treated with idelalisib (28–30). To assess
the clinical potential of a kinase inhibitor for CLL therapy, the combined on- and off-target effects on both malignant and healthy cells need to be taken into account. In this study, we perform comparative studies of the effects of JAK, BTK, and PI3Kd in- hibitors on CLL cells and healthy immune cells. We determine both beneficial and detrimental effects of all kinase inhibitors and ex- plore whether therapeutic rationales exist to use combinations of these inhibitors not only for CLL therapy but also for other diseases.

Materials and Methods
Patient and healthy donor samples
Peripheral blood samples from untreated CLL patients were collected at the Amsterdam University Medical Centers (Amsterdam UMC), Aca- demic Medical Center in Amsterdam, the Netherlands. Healthy donor PBMC were isolated from buffy coats obtained from Sanquin Blood Supply (Amsterdam, the Netherlands). Ethical approval was provided by the medical ethical committee at the Amsterdam UMC, Academic Medical Center, and written informed consent was obtained in accordance with the Declaration of Helsinki. PBMC from CLL patients and healthy donors were isolated and cryopreserved as de- scribed earlier (31).
Cell lines
NIH-3T3 fibroblasts (ACC number 59; Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany) and 3T40L (NIH-3T3 fibroblasts expressing CD40L) were cultured in IMDM (Thermo Fisher Scientific, Waltham, MA) supplemented with 10% FCS and 1% penicillin/streptomycin.
Compounds
The JAK inhibitors momelotinib and ruxolitinib, the BTK inhibitor tirabrutinib, and the PI3Kd inhibitor idelalisib were all obtained from Gilead Sciences (Foster City, CA). The BTK inhibitor ibrutinib was purchased from Selleckchem (Houston, TX). Venetoclax (ABT-199) was purchased from Sanbio (Uden, the Netherlands). Fludarabine was purchased from Sigma- Aldrich (St Louis, MO). All compounds were used in concentrations close to their EC50 values or clinically relevant levels.
Proliferation and chemoresistance of CLL cells
For proliferation experiments, PBMC from CLL patients were labeled with
0.5 mM CFSE (Thermo Fisher Scientific) and cultured for 5 d on either NIH-3T3 or 3T40L cells with or without 25 ng/ml rIL-21 (Invitrogen, Carlsbad, CA). Cultures were measured on a FACSCanto (BD Biosciences, San Jose, CA), and data were analyzed using FlowJo Version 10. To study chemoresistance of CLL cells, PBMC from CLL patients were cocultured with NIH-3T3 or 3T40L cells for 3 d. Cells were treated with venetoclax or fludarabine for 24 h. Target cell death was analyzed by incubating cultures with DiOC6 (Thermo Fisher Scientific) and propidium iodide (PI) (Sigma- Aldrich). Cells were analyzed on a FACSCanto flow cytometer. Data were analyzed using FlowJo Version 10.
CLL cell viability and Western blot analysis
CLL-derived PBMC were preincubated for 1 h with inhibitors and stim- ulated for 24 h with IL-4 (10 ng/ml; Bio-Techne, Minneapolis, MN) and during the last 30 min with anti-IgM (20 mg/ml; BioLegend, San Diego, CA). Viability was measured by FACS using DiOC6/PI staining as de- scribed above. For Western blot, cell lysates were prepared by lysing in RIPA buffer (150 mM NaCl, 1 mM EDTA, 50 mM Tris–HCl [pH 7.4], 0.1% SDS, and 1% NP-40) and, subsequently, subjected to 10 s of soni- cation in a Branson sonicator (Danbury, CT). Lysates were analyzed by SDS-PAGE. Western blot was performed using the following Abs: rabbit anti–p-STAT6, rabbit anti-pErk, rabbit anti-pAkt (Ser473), rabbit anti-pS6, mouse anti-S6 (all from Cell Signaling Technology [Danvers, MA]) and goat anti-actin (Santa Cruz Biotechnology, Dallas, TX).
Proliferation, IgM/IgG production and differentiation of healthy B cells
Healthy donor PBMC were stained with CFSE, and stimulated with CpG ODN2006 (1 mg/ml; Invitrogen) and IL-2 (100 U/ml) for 6 d. Fresh drugs were added after 3 d of stimulation. Cells were stained with the following Abs for flow cytometry: CD19-PerCP-Cy5.5, CD20- allophycocyanin-H7, IgD–PE, CD27–allophycocyanin, and CD38-PE-Cy7 (all from BD

Biosciences). Samples were measured on a FACSCanto flow cytometer. IgM and IgG levels were measured by ELISA in culture supernatants, as was described earlier (32), using polyclonal rabbit anti-human IgG and IgM reagents and a serum protein calibrator (all from Agilent Technolo- gies, Santa Clara, CA).
NK cell proliferation, cytokine production, and cytotoxicity
Healthy donor PBMC were stimulated overnight with either a combination of IL-2 (100 U/ml; PeproTech, Rocky Hill, NJ) and IL-15 (10 ng/ml; PeproTech) or a combination of IL-2, IL-12 (10 mg/ml; R&D Systems, Minneapolis, MN), and IL-18 (100 mg/ml; R&D Systems). PBMC frac- tions were cocultured with K562 target cells (American Type Culture Collection, Manassas, VA) for 4 h to stimulate NK cells. Cells were stained with the following Abs for flow cytometry: CD107a-PE-Cy7, CD56– BUV395, CD3–V500, CD16–BV450, IFN-g–BV421, TNF-a–BV650, and
granzyme B–Alexa Fluor 700 (all BD from Biosciences) and LIVE/DEAD Fixable Red Stain (Invitrogen). Intracellular stainings were performed using Cytofix/Cytoperm reagent kit according to the manufacturer’s pro- tocol (BD Biosciences). For NK cell cytotoxicity, stimulated PBMC were cocultured with CellTrace Violet (Invitrogen)–labeled K562 target cells for 3 h in a NK:K562 at a ratio of 1:1. Cell cultures were labeled with MitoTracker Orange (Invitrogen) and TO-PRO-3 (Invitrogen) to determine target cell death. For NK cell proliferation, PBMC were labeled with CFSE as described above. PBMC were stimulated with IL-2 plus IL-15 or IL-2 plus IL-12 plus IL-18 for 5 d in the concentrations described above. Fresh drugs were added after 3 d of stimulation. Samples were analyzed on an LSRFortessa (BD Biosciences), and data were analyzed using FlowJo Version 10.
T cell stimulation using anti-CD3 and anti-CD28 Abs
Healthy donor PBMC were labeled with CFSE and stimulated with anti- CD3 (clone 1XE) and anti-CD28 (clone 15E8) Abs for 4 d. Afterwards, cells were stained with CD3–Alexa Fluor 700 (Thermo Fisher Scientific), CD4-PE-Cy7, CD8-PerCP-Cy5.5, CD25–allophycocyanin (all BD Bio- sciences), and LIVE/DEAD Fixable Red Stain (Invitrogen). Samples were measured on a FACSCanto flow cytometer, and data were analyzed using FlowJo Version 10. IFN-g production was measured in the culture super- natant using a human IFN-g Uncoated ELISA Kit, according to manufac- turer’s protocol (Thermo Fisher Scientific). Pure fractions of CD4 and CD8 T cells were isolated by positive selection with MACS magnetic beads (Miltenyi Biotec, Bergish Gladbach, Germany) according to manufacturer’s instruction.
Mixed lymphocyte reaction
CD14+ cells were isolated by positive selection with MACS beads (Mil- tenyi Biotec) and differentiated into monocyte-derived dendritic cells (moDCs) by incubation with IL-4 (20 ng/ml; R&D Systems) and GM-CSF (1000 U/ml; Genzyme, Cambridge, MA) for 7 d and maturated with LPS (100 ng/ml; Sigma-Aldrich) during the last 2 d. Healthy donor PBMC were labeled with CFSE and incubated with allogeneic moDCs for 4 d. After- wards, analysis of T cell proliferation, activation, and IFN-g production by ELISA was performed as described above.
Statistics
To visualize the relative effect of drug treatments on immune cell function, data of all experiments were normalized to the stimulated control without any drug added. Statistical analysis was performed on raw data before normali- zation. Data were analyzed using repeated measures one-way ANOVA followed by a Dunnett multiple comparisons test. Statistical analysis was per- formed using GraphPad Prism v7. Differences between groups were considered significant when p , 0.05.

Results
JAK inhibitors block proliferation induced by CD40/IL-21 but do not induce cell death in CLL cells
Within the microenvironment, CLL cells receive a variety of signals, including through cytokines, BCR activation, and costimulation via TNFR. The effects of kinase inhibition on microenvironmental stimulation was tested using the JAK in- hibitors momelotinib and ruxolitinib as well as the BTK in- hibitor ibrutinib and the PI3Kd inhibitor idelalisib, currently used to treat CLL patients. Tirabrutinib (GS4059) is a newly developed BTK inhibitor that is more selective compared with

FIGURE 1. Effects of JAK inhibitors on CLL cells within the tumor microenvironment. The effect of JAK inhibitors on CLL signaling pathways within the tumor microenvironment. (A) CLL PBMC were cocultured on 3T40L cells and treated with IL-21 for 5 d to induce CLL cell proliferation in the presence of either JAK, BTK, or PI3Kd inhibitors (n = 8). (B) CLL PBMC were stimulated with IL-4 (24 h) and anti-IgM (30 min) or medium control, with or without kinase inhibitors (n = 3). Cell viability was measured by incubating cells with DiOC6 and PI (live cells defined as DiOC6+PI2). (C) CLL PBMC were cultured on 3T3 or 3T40L fibroblasts for 3 d in combination with JAK inhibitors. Afterwards, cells were treated for 24 h with either venetoclax or fludarabine (n = 3). Viability was measured by incubating cells with DiOC6 and PI. (D) CLL cells were stimulated with IL-4 (Figure legend continues)
ibrutinib (33, 34). Proliferation of CLL cells was induced by coculturing primary CLL cells on CD40L-expressing 3T3 fibro- blasts in combination with IL-21 (12). Momelotinib was able to block IL-21 signaling and reduce proliferation of CLL cells to a greater extent than ruxolitinib (Fig. 1A, Supplemental Fig. 1A–C). As we have shown before, CD40L/IL-21–induced proliferation can be partially inhibited by the BTK inhibitors ibrutinib and the PI3Kd inhibitor idelalisib (19), and this also held for the BTK inhibitor tirabrutinib. Momelotinib showed the strongest inhibition of proliferation of CLL cells. JAK, BTK, and PI3Kd inhibitors did not induce cell death in CLL cells (Fig. 1B). Cell death of CLL cells can be induced by other therapeutic agents, such as the Bcl-
2 inhibitor venetoclax or the chemotherapeutic agent fludar- abine. We have previously demonstrated that CD40 stimulation, as a model for lymph node prosurvival signals, renders CLL cells resistant to both venetoclax and fludarabine (35, 36). As ex- pected, JAK inhibitors were not able to reduce resistance to venetoclax and fludarabine after coculture of CLL cells on CD40L-fibroblasts because CD40 signaling is not mediated by JAKs (Fig. 1C). To study the effects of JAK inhibitors within the CLL microenvironment compared with other kinase inhibitors, we studied their effects on signaling by IL-4 and IgM [a model for both T cell help and BCR stimulation within the lymph node (11)] by Western blot. Treatment with both JAK inhibitors but not BTK or PI3Kd inhibitors led to a reduction in p-STAT6 in- duced by IL-4 (Fig. 1D). Surprisingly, momelotinib treatment also led to a reduction in IgM-induced p-Akt and p-S6 levels, although not as strong as both BTK inhibitors or idelalisib. Ruxolitinib did not affect BCR signaling to Akt or S6, demon- strating different modes of action of these JAK inhibitors. These results demonstrate that JAK inhibitors are not cytotoxic for CLL cells by themselves, but are able to influence signaling of pro- survival cytokines like IL-4 and IL-21 that induce proliferation and IgM expression, and momelotinib was able to partially block BCR signaling.

JAK inhibition minimally affects healthy B cell function
New small molecules used for CLL therapy can affect healthy B cells as well, especially the BTK inhibitors. Proliferation of healthy B cells by stimulation with CpG/IL-2 was not inhibited by JAK inhibitors (Fig. 2A, Supplemental Fig. 2A–C). In contrast, BTK and PI3Kd inhibitors were able to significantly reduce proliferation of healthy B cells. Similar to CLL B cells, JAK, BTK, and PI3Kd inhibition did have an effect on proliferation of healthy B cells induced by CD40L and IL-21 stimulation (Supplemental Fig. 2D). None of the kinase inhibitors signifi- cantly affected IgM production after CpG/IL-2, although ibrutinib and idelalisib showed a clear trend of inhibition (Fig. 2B). A high dose of momelotinib was able to inhibit IgG production, in contrast to ruxolitinib. As expected, both BTK inhibitors and idelalisib affected IgG production (Fig. 2C). Finally, differentiation of healthy B cells in response to CpG/ IL-2 was not significantly altered by JAK inhibitors (Fig. 2D). BTK inhibition led to a slight reduction of B cell differentia- tion, whereas PI3Kd inhibition had no effect. These results indicate that JAK inhibition has only a moderate effect on the function of healthy B cells, in contrast to BTK and PI3Kd inhibi- tors, which inhibit the function of healthy B cells to a greater extent.

JAK inhibition ablates cytokine priming of NK cells but not the activity via natural cytotoxicity receptors
Although NK cells do not require stimulation via cytokines, cy- tokines can increase the magnitude of NK cell responses (37). PBMC were stimulated with either a combination of IL-2/IL-15 or IL-2/IL-12/IL-18 to activate NK cells and, subsequently, in- cubated with the classical NK target cell line K562. IL-2/IL-15 stimulation induced NK cell proliferation and increased effector responses of NK cells toward K562 cells, resulting in increased production of IFN-g, TNF-a, and higher levels of degranulation and target cell death (Fig. 3A–E, Supplemental Fig. 3A–C). Because both IL-2 and IL-15 signaling are dependent on JAK signaling, activation via these cytokines was efficiently blocked by both JAK inhibitors, leading to a reduction of NK responses. Ibrutinib also showed a strong inhibitory effect on NK cell function after IL-2/IL-15 stimulation, especially on the produc- tion of effector cytokines, whereas tirabrutinib and idelalisib had smaller effects (Fig. 3A–E). Stimulation with IL-2/IL-12/IL-18 also enhanced effector responses of NK cells (Supplemental Fig. 3D–H). NK cell proliferation upon IL-2/IL-12/IL-18 was blocked completely by JAK inhibitors, whereas no or only partial effects on IL-2/IL-12/IL-18–enhanced cytokine production and cytotoxicity were observed. Because IL-18 signaling is not JAK dependent but is a TLR-like stimulus, it can be assumed that this stimulating signal is not affected by JAK inhibition, and NK cells continue to receive an activating signal in this setting. Both BTK inhibitors and idelalisib had only small effects on IL-2/IL-12/IL-18 stimulation, although ibrutinib substantially affected NK cell proliferation and TNF-a production. Importantly, NK cell cytokine production and cytotoxicity in response to target cells was not completely abrogated by JAK inhibitors but returned to levels similar to untreated PBMC, indicating that signaling via JAK-independent natural cytotoxicity receptors is still functioning and that JAK inhibition only targets the cytokine signaling pathways.
JAK inhibition targets T cell activation and cytokine production
JAKs are involved in cytokine-induced amplification of T cell re- sponses and differentiation of specific T cell phenotypes. Stimulation of PBMC with anti-CD3 and anti-CD28 Abs induced prolifera- tion, activation, IFN-g production, and T cell differentiation in both CD4 and CD8 T cells (Fig. 4A–F, Supplemental Fig. 4A–D). Interestingly, ibrutinib showed the strongest inhibition of T cell proliferation and activation. This was probably mediated via off- target inhibition of ITK (26, 27), as the more selective tirabrutinib consistently showed less off-target effects on T cell function at any level. JAK inhibitors showed modest inhibition of CD8 T cell proliferation and activation, yet the production of IFN-g was strongly inhibited, especially by momelotinib (Fig. 4C, 4D). Although results did not reach statistical significance because of patient variability, a clear relative effect within donors was ob- served (Fig. 4D). The inhibitory effects on IFN-g production of momelotinib, ruxolitinib, ibrutinib, and idelalisib are a direct effect on CD4 and CD8 T cells, as IFN-g production by pu- rified CD4 and CD8 T cell fractions was also inhibited (Fig. 4E). Similar to the data in full PBMC, CD8 T cells seemed more sensitive to JAK inhibitors compared with CD4 T cells. Although idelalisib showed minimal effects on T cell activation and proliferation, it significantly affected T cell differentiation and IgM in the presence of JAK, BTK, or PI3Kd inhibitors. Representative Western blot and quantification showing phosphorylation of target molecules of IL-4 and IgM signaling. LY294002 (1 mM) and rapamycin (1 mM) were used as controls for PI3K and mTOR signaling, respectively. Bar graphs show summarized relative protein expression of three independent experiments. Bars indicate mean 6 SD relative to condition without inhibitor. *p , 0.05,**p , 0.01, repeated measures one-way ANOVA followed by Dunnett multiple comparisons test (statistics were performed on nontransformed data).

FIGURE 2. Effect of kinase inhibitors on healthy B cell function. Healthy donor PBMC were stimulated with CpG and IL-2 for 6 d in combination with JAK, BTK, and PI3Kd inhibitors (n = 8 for all experiments). (A) Proliferation of B cells after 6 d culture. (B) Levels of secreted IgM in culture supernatant measured by ELISA. (C) Levels of secreted IgG in culture supernatant measured by ELISA (D) Subset differentiation of B cells 6 d after stimulation. Representative examples of subset marker expression is shown in contour plots, quantification of multiple experiments are (Figure legend continues)

FIGURE 3. Effect of JAK inhibitors on NK cell function. Healthy donor PBMC were stimulated with IL-2 and IL-15 for 5 d (A) or overnight (B–E) in combination with kinase inhibitors (n = 8). (A) Proliferation of NK cells after stimulation for 5 d. (B and C) Percentage of NK cells producing IFN-g (B) or TNF-a (C) after 4 h coculture of stimulated PBMC with K562 target cells, as measured by flow cytometry. (D) Percentage of degranulated (CD107a+) NK cells after 4 h coculture with K562 target cells. (E) Specific lysis of K562 target cells after coculture with stimulated PBMC for 3 h. Bars indicate mean 6 SD relative to condition without inhibitor. *p , 0.05, **p , 0.01, ***p , 0.001, ****p , 0.0001, repeated measures one-way ANOVA followed by Dunnett multiple comparisons test (statistics were performed on nontransformed data).
after stimulation, leading to an increase in effector cell differ- entiation. Conversely, ibrutinib had a small inhibitory effect on T cell differentiation (Fig. 4F), whereas JAK inhibition had no effect (Supplemental Fig. 4C).
JAK inhibitors and ibrutinib strongly affect allogeneic T cell responses
Because both JAK inhibitors and ibrutinib have clinical activity in GvHD, we also tested the effects of the inhibitors in a mixed lymphocyte reaction. PBMC from healthy donors were mixed with allogeneic LPS-maturated moDC, and T cell proliferation, activa- tion, and cytokine production were determined after 4 d of coculture. Allogeneic T cell responses were significantly inhibited by both

JAK inhibitors (especially momelotinib) and ibrutinib (Fig. 5A, 5B). Tirabrutinib showed no inhibition of T cell function, in sharp con- trast to ibrutinib and in accordance with the response to stimulation with anti-CD3 and anti-CD28 Abs. The inhibition of T cell activa- tion and proliferation by JAK inhibitors corresponded with a dose- dependent decrease in IFN-g levels (Fig. 5C). In contrast, IFN-g levels consistently increased using the lower dose of ibrutinib, and a similar trend was seen with tirabrutinib.

Discussion
In the context of increasing application of kinase inhibitors in can- cer treatment and emerging awareness of infectious complications of
shown in bar graphs. B cell subsets are defined as follows: naive (IgD+CD272), memory (IgD2CD27+), and plasmablast (CD27++CD38+). Bars indicate mean 6 SD relative to condition without inhibitor. *p , 0.05, **p , 0.01, repeated measures one-way ANOVA followed by Dunnett multiple comparisons test (statistics were performed on nontransformed data).

FIGURE 4. Effect of kinase inhibitors on the function of healthy T cells. Healthy donor PBMC were stimulated with anti-CD3 and anti-CD28 Abs for 4 d in combination with JAK, BTK, and PI3Kd inhibitors (n = 8). (A) Proliferation of CD4 and CD8 T cells after stimulation for 4 d. (B) Expression of activation marker CD25 on the surface of CD4 and CD8 T cells after 4 d of stimulation. (C) Levels of IFN-g in culture supernatants measured by ELISA. (D) Normalized values of (C), showing a clear relative effect of most kinase inhibitors on IFN-g production of T cells. (E) IFN-g production of MACS-purified CD4 and CD8 T cell fractions after 4 d of stimulation (n = 4). (F) T cell subset differentiation after 4 d of stimulation. T cell subsets are defined as follows: naive (CD45RA+CD27+), memory (CD45RA2CD27+), effector (CD45RA2CD272), and RA-expressing effector memory (EMRA) (CD45RA+CD272). Bars indicate mean 6 SD relative to condition without inhibitor, except for (C), which depicts nontransformed data. *p , 0.05, **p , 0.01, repeated measures one-way ANOVA followed by Dunnett multiple comparisons test (statistics were performed on nontransformed data).these compounds, a thorough understanding of the effects of JAK, BTK, and PI3Kd inhibitors on the function of both healthy and malignant lymphocytes is warranted. JAK inhibitors could play a beneficial role for the treatment of CLL by blocking signaling of important prosurvival molecules like IL-4 and IL-21. Although our

results indicate that JAK inhibitors do not have detrimental effects on the function of healthy B cells (in contrast to BTK and PI3Kd in- hibitors), the observed inhibition of NK and T cell function could have significant side effects during patient treatment. Off-target ef- fects on T cell function also occurs with ibrutinib, leading to

FIGURE 5. Kinase inhibitors modulate allogeneic T cell responses. Healthy donor PBMC were stimulated with allogeneic LPS-maturated moDCs for 4 d in combination with JAK, BTK, and PI3Kd inhibitors (n = 8). (A) Proliferation of CD4 and CD8 T cells after coculture for 4 d. (B) Expression of activation marker CD25 on the surface of CD4 and CD8 T cells after 4 d of coculture. (C) Levels of IFN-g in culture supernatants measured by ELISA. Bars indicate mean 6 SD relative to condition without inhibitor. *p , 0.05, **p , 0.01, ***p , 0.001, ****p , 0.0001, repeated measures one-way ANOVA followed by Dunnett multiple comparisons test (statistics were performed on nontransformed data).
functional impairments, but these effects are not observed with the more selective BTK inhibitor tirabrutinib.
We studied the effect of kinase inhibitors on healthy and ma- lignant B cell function. Although momelotinib and ruxolitinib are both JAK1/2 inhibitors, we observe clear differential effects on BCR signaling. In particular, momelotinib inhibits BCR-mediated phosphorylation of Akt and S6, whereas ruxolitinib did not target BCR signaling. The inhibition of BCR signaling by momelotinib might be beneficial during CLL therapy, as the BCR pathway plays an important role in the pathology of the disease (11). Recently, a role for JAK2 has been described in BTK activation, especially in

the context of CXCR4 signaling (38). If the effects of momelotinib on BCR signaling are via on-target JAK2 inhibition, it is unclear why ruxolitinib does not induce similar effects. Therefore, differ- ential off-target effects of momelotinib and ruxolitinib might ex-
plain our observations. Momelotinib also inhibits TANK-binding kinase 1 (TBK1) and IKKε, kinases that are involved in NF-kB signaling upon activation by TLR and RIG1 signaling, suggesting
that these kinases play a role in the inhibitory effect of momelotinib on BCR signaling (39).
Off-target effects of kinase inhibitors have been described [e.g., the observed effect of ibrutinib on T cell function is in line with
reports on off-target ITK inhibition (26, 27)]. More specific BTK inhibitors, such as tirabrutinib, have been developed to limit off- target effects (33, 34). Tirabrutinib showed similar inhibition of B cell function in both healthy B cells and CLL cells but no in- hibition of T cell function. Allogeneic T cell responses were also inhibited by ibrutinib but not tirabrutinib. Surprisingly, IFN-g levels increased in mixed lymphocyte reactions with both inhibi- tors, suggesting a BTK-mediated effect that is perhaps explained by BTK-dependent improvement of the Ag-presenting capacity of dendritic cells (40). The disappearance of this effect with a higher dose of ibrutinib may well reflect ITK inhibition.
Tirabrutinib may, therefore, have less infectious complications compared with ibrutinib. Together with promising phase 1 trial results (33, 34), these data support further exploration of tirabrutinib or other more selective BTK inhibitors as a clinical treatment in CLL and other malignancies. In contrast, off-target effects of ibrutinib on T cells are not always detrimental, as has been observed in CLL, in which ibrutinib treatment increases T cell numbers and improves the efficacy of chimeric AgR T cell therapy (27, 41).
Several prosurvival stimuli that CLL cells receive from the microenvironment are mediated via JAK signaling, like T cell– derived IL-4 and IL-21 (11, 12). In CLL patients, ruxolitinib leads to lymphocyte redistribution out of the lymph nodes and disease stabilization, establishing in vivo activity of JAK inhibition (20, 21). Although the efficacy of ruxolitinib monotherapy in patients unfit for regular chemo-immunotherapy is only moderate, the egress of tumor cells from lymph nodes observed with ruxolitinib treatment is similar to the lymphocytosis seen during the early phases of ibrutinib or idelalisib treatment (20, 21). JAK inhibi- tors may therefore be useful compounds in combination with cell death–inducing agents, such as the Bcl-2 inhibitor venetoclax, even in patients refractory to ibrutinib or idelalisib, by depriving tumor cells of microenvironmental stimuli. Because combination treatment with venetoclax and ibrutinib has shown promising clinical efficacy in CLL (42, 43), combinatory treatment of JAK inhibitors and venetoclax may pose similar clinical benefit.
The strong inhibitory effects on T and NK cell function by JAK inhibitors observed by us and others (23, 24) may complicate their use in CLL. Ruxolitinib treatment is associated with an increased risk of infectious complications, predominantly with infections of the urogenital and respiratory tract, but reactivation of hepatitis B and tuberculosis also occurs (44). The increased risk of infections may pose a significant problem for already frail CLL patients, where infections are a leading cause of death (1). Indeed, rux- olitinib treatment led to infectious complications in unfit CLL patients (21).
The functional impairment of lymphocytes by JAK inhibitors can be beneficial in other disease settings, such as autoimmune diseases and GvHD. Both T and B cells are implicated in GvHD pathology (45), and ruxolitinib has clinical efficacy in refractory GvHD patients (46). However, we show, in this study, that JAK inhibitors have relatively mild effects on B cell function, which may indicate that pathogenic B cell responses remain relatively intact in GvHD during ruxolitinib treatment. The inhibition of both T and B cell function provides a rationale to use ibrutinib for GvHD therapy. Indeed, our data demonstrate that ibrutinib also inhibits T cells responses toward allogeneic cells. In clinical trials, ibrutinib treatment led to clinical responses in two-thirds of refractory GvHD patients, with efficacy in all affected organs and a reduction in steroid use, and ibrutinib is now a Food and Drug Administration–approved drug for glucocorticoid-resistant GvHD (47).
In conclusion, we show that JAK inhibitors potently inhibit several prosurvival stimuli for CLL cells. However, their inhibition

of T and NK cell function, and, consequently, the increased risk of infections, may complicate clinical treatment with JAK inhib- itors in CLL patients. The BTK inhibitor tirabrutinib consistently showed a similar inhibitory potential as ibrutinib but lacked the off- target effects of ibrutinib on T cells, which warrants future research comparing tirabrutinib with ibrutinib in the clinical setting. Con- versely, the inhibition of T and NK cell function we observe by JAK inhibitors and ibrutinib can be beneficial in disease settings with unwanted lymphocyte activation, such as GvHD. Our data are in line with early clinical data and demonstrate that kinase inhibitors in development as antitumor drugs in hematological malignancies can also be applied to block unwanted lymphocyte function in other diseases. The properties of individual kinase inhibitors can be exploited in combination treatment strategies tailored to the in- hibitory effects desired.

Disclosures
S.T. was an employee of Gilead Sciences during this investigation. This re- search was sponsored via a research agreement between Amsterdam UMC and Gilead Sciences.
References
1. Kipps, T. J., F. K. Stevenson, C. J. Wu, C. M. Croce, G. Packham, W. G. Wierda,
S. O’Brien, J. Gribben, and K. Rai. 2017. Chronic lymphocytic leukaemia. Nat. Rev. Dis. Primers 3: 16096.
2. Byrd, J. C., J. R. Brown, S. O’Brien, J. C. Barrientos, N. E. Kay, N. M. Reddy,
S. Coutre, C. S. Tam, S. P. Mulligan, U. Jaeger, et al; RESONATE Investigators. 2014. Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N. Engl. J. Med. 371: 213–223.
3. Roberts, A. W., M. S. Davids, J. M. Pagel, B. S. Kahl, S. D. Puvvada, J. F. Gerecitano,
T. J. Kipps, M. A. Anderson, J. R. Brown, L. Gressick, et al. 2016. Targeting BCL2 with venetoclax in relapsed chronic lymphocytic leukemia. N. Engl. J. Med. 374: 311–322.
4. Woyach, J. A., R. R. Furman, T. M. Liu, H. G. Ozer, M. Zapatka, A. S. Ruppert,
L. Xue, D. H. Li, S. M. Steggerda, M. Versele, et al. 2014. Resistance mecha- nisms for the Bruton’s tyrosine kinase inhibitor ibrutinib. N. Engl. J. Med. 370: 2286–2294.
5. Woyach, J. A., and A. J. Johnson. 2015. Targeted therapies in CLL: mechanisms of resistance and strategies for management. Blood 126: 471–477.
6. Herling, C. D., N. Abedpour, J. Weiss, A. Schmitt, R. D. Jachimowicz,
O. Merkel, M. Cartolano, S. Oberbeck, P. Mayer, V. Berg, et al. 2018. Clonal dynamics towards the development of venetoclax resistance in chronic lym- phocytic leukemia. Nat. Commun. 9: 727.
7. Blombery, P., M. A. Anderson, J. N. Gong, R. Thijssen, R. W. Birkinshaw,
E. R. Thompson, C. E. Teh, T. Nguyen, Z. Xu, C. Flensburg, et al. 2019. Ac- quisition of the recurrent Gly101Val mutation in BCL2 confers resistance to venetoclax in patients with progressive chronic lymphocytic leukemia. Cancer Discov. 9: 342–353.
8. Forconi, F., and P. Moss. 2015. Perturbation of the normal immune system in patients with CLL. Blood 126: 573–581.
9. Hamblin, A. D., and T. J. Hamblin. 2008. The immunodeficiency of chronic lymphocytic leukaemia. Br. Med. Bull. 87: 49–62.
10. van Attekum, M. H., E. Eldering, and A. P. Kater. 2017. Chronic lymphocytic leukemia cells are active participants in microenvironmental cross-talk. Haematologica 102: 1469–1476.
11. Aguilar-Hernandez, M. M., M. D. Blunt, R. Dobson, A. Yeomans,
S. Thirdborough, M. Larrayoz, L. D. Smith, A. Linley, J. C. Strefford, A. Davies, et al. 2016. IL-4 enhances expression and function of surface IgM in CLL cells. Blood 127: 3015–3025.
12. Pascutti, M. F., M. Jak, J. M. Tromp, I. A. Derks, E. B. Remmerswaal,
R. Thijssen, M. H. van Attekum, G. G. van Bochove, D. M. Luijks, S. T. Pals, et al. 2013. IL-21 and CD40L signals from autologous T cells can induce antigen-independent proliferation of CLL cells. Blood 122: 3010–3019.
13. Harrison, C., J. J. Kiladjian, H. K. Al-Ali, H. Gisslinger, R. Waltzman,
V. Stalbovskaya, M. McQuitty, D. S. Hunter, R. Levy, L. Knoops, et al. 2012. JAK inhibition with ruxolitinib versus best available therapy for myelofibrosis. N. Engl. J. Med. 366: 787–798.
14. Verstovsek, S., R. A. Mesa, J. Gotlib, R. S. Levy, V. Gupta, J. F. DiPersio,
J. V. Catalano, M. Deininger, C. Miller, R. T. Silver, et al. 2012. A double-blind, placebo-controlled trial of ruxolitinib for myelofibrosis. N. Engl. J. Med. 366: 799–807.
15. Pardanani, A., R. R. Laborde, T. L. Lasho, C. Finke, K. Begna, A. Al-Kali,
W. J. Hogan, M. R. Litzow, A. Leontovich, M. Kowalski, and A. Tefferi. 2013. Safety and efficacy of CYT387, a JAK1 and JAK2 inhibitor, in myelofibrosis. Leukemia 27: 1322–1327.
16. Mesa, R. A., J. J. Kiladjian, J. V. Catalano, T. Devos, M. Egyed, A. Hellmann,
D. McLornan, K. Shimoda, E. F. Winton, W. Deng, et al. 2017. SIMPLIFY-1: a phase III randomized trial of momelotinib versus ruxolitinib in Janus kinase inhibitor-na¨ıve patients with myelofibrosis. J. Clin. Oncol. 35: 3844–3850.
17. Modi, B., M. Hernandez-Henderson, D. Yang, J. Klein, S. Dadwal, E. Kopp,
K. Huelsman, S. Mokhtari, H. Ali, M. M. A. Malki, et al. 2019. Ruxolitinib as salvage therapy for chronic graft-versus-host disease. Biol. Blood Marrow Transplant. 25: 265–269.
18. Khoury, H. J., A. A. Langston, V. K. Kota, J. A. Wilkinson, I. Pusic, A. Jillella,
S. Bauer, A. S. Kim, D. Roberts, Z. Al-Kadhimi, et al. 2018. Ruxolitinib: a steroid sparing agent in chronic graft-versus-host disease. Bone Marrow Transplant. 53: 826–831.
19. Slinger, E., R. Thijssen, A. P. Kater, and E. Eldering. 2017. Targeting antigen- independent proliferation in chronic lymphocytic leukemia through differential kinase inhibition. Leukemia 31: 2601–2607.
20. Jain, P., M. Keating, S. Renner, C. Cleeland, H. Xuelin, G. N. Gonzalez,
D. Harris, P. Li, Z. Liu, I. Veletic, et al. 2017. Ruxolitinib for symptom control in patients with chronic lymphocytic leukaemia: a single-group, phase 2 trial. Lancet Haematol. 4: e67–e74.
21. Spaner, D. E., G. Wang, L. McCaw, Y. Li, P. Disperati, M. A. Cussen, and Y. Shi. 2016. Activity of the Janus kinase inhibitor ruxolitinib in chronic lymphocytic leukemia: results of a phase II trial. Haematologica 101: e192–e195.
22. McLornan, D. P., A. A. Khan, and C. N. Harrison. 2015. Immunological conse- quences of JAK inhibition: friend or foe? Curr. Hematol. Malig. Rep. 10: 370–379.
23. Perner, F., T. M. Schno¨ der, S. Ranjan, D. Wolleschak, C. Ebert, M. C. Pils,
S. Frey, A. Polanetzki, C. Fahldieck, U. Scho¨nborn, et al. 2016. Specificity of JAK-kinase inhibition determines impact on human and murine T-cell function. Leukemia 30: 991–995.
24. Scho¨nberg, K., J. Rudolph, M. Vonnahme, S. Parampalli Yajnanarayana,
I. Cornez, M. Hejazi, A. R. Manser, M. Uhrberg, W. Verbeek, S. Koschmieder, et al. 2015. JAK inhibition impairs NK cell function in myeloproliferative neoplasms. Cancer Res. 75: 2187–2199.
25. Heine, A., S. A. Held, S. N. Daecke, S. Wallner, S. P. Yajnanarayana, C. Kurts,
D. Wolf, and P. Brossart. 2013. The JAK-inhibitor ruxolitinib impairs dendritic cell function in vitro and in vivo. Blood 122: 1192–1202.
26. Dubovsky, J. A., K. A. Beckwith, G. Natarajan, J. A. Woyach, S. Jaglowski,
Y. Zhong, J. D. Hessler, T. M. Liu, B. Y. Chang, K. M. Larkin, et al. 2013. Ibrutinib is an irreversible molecular inhibitor of ITK driving a Th1-selective pressure in T lymphocytes. Blood 122: 2539–2549.
27. Long, M., K. Beckwith, P. Do, B. L. Mundy, A. Gordon, A. M. Lehman,
K. J. Maddocks, C. Cheney, J. A. Jones, J. M. Flynn, et al. 2017. Ibrutinib treatment improves T cell number and function in CLL patients. J. Clin. Invest. 127: 3052– 3064.
28. Martinelli, S., R. Maffei, S. Fiorcari, C. Quadrelli, P. Zucchini, S. Benatti,
L. Potenza, M. Luppi, and R. Marasca. 2018. Idelalisib impairs T-cell-mediated immunity in chronic lymphocytic leukemia. Haematologica 103: e598–e601.
29. Dong, S., B. K. Harrington, E. Y. Hu, J. T. Greene, A. M. Lehman, M. Tran,
R. L. Wasmuth, M. Long, N. Muthusamy, J. R. Brown, et al. 2019. PI3K p110d inactivation antagonizes chronic lymphocytic leukemia and reverses T cell im- mune suppression. J. Clin. Invest. 129: 122–136.
30. Hanna, B. S., P. M. Roessner, A. Scheffold, B. M. C. Jebaraj, Y. Demerdash,
S. O¨ ztu¨rk, P. Lichter, S. Stilgenbauer, and M. Seiffert. 2019. PI3Kd inhibition modulates regulatory and effector T-cell differentiation and function in chronic
lymphocytic leukemia. Leukemia 33: 1427–1438.
31. Mackus, W. J., F. N. Frakking, A. Grummels, L. E. Gamadia, G. J. De Bree,
D. Hamann, R. A. Van Lier, and M. H. Van Oers. 2003. Expansion of CMV- specific CD8+CD45RA+CD27- T cells in B-cell chronic lymphocytic leukemia. Blood 102: 1057–1063.
32. Kuijpers, T. W., R. J. Bende, P. A. Baars, A. Grummels, I. A. Derks,
K. M. Dolman, T. Beaumont, T. F. Tedder, C. J. van Noesel, E. Eldering, and
R. A. van Lier. 2010. CD20 deficiency in humans results in impaired T cell- independent antibody responses. J. Clin. Invest. 120: 214–222.

33. Walter, H. S., S. A. Rule, M. J. Dyer, L. Karlin, C. Jones, B. Cazin, P. Quittet,
N. Shah, C. V. Hutchinson, H. Honda, et al. 2016. A phase 1 clinical trial of the selective BTK inhibitor ONO/GS-4059 in relapsed and refractory mature B-cell malignancies. Blood 127: 411–419.
34. Walter, H. S., S. Jayne, S. A. Rule, G. Cartron, F. Morschhauser, S. Macip,
L. Karlin, C. Jones, C. Herbaux, P. Quittet, et al. 2017. Long-term follow-up of patients with CLL treated with the selective Bruton’s tyrosine kinase inhibitor ONO/GS-4059. Blood 129: 2808–2810.
35. Thijssen, R., J. Ter Burg, G. G. van Bochove, M. F. de Rooij, A. Kuil,
M. H. Jansen, T. W. Kuijpers, J. W. Baars, A. Virone-Oddos,
M. Spaargaren, et al. 2016. The pan phosphoinositide 3-kinase/mammalian target of rapamycin inhibitor SAR245409 (voxtalisib/XL765) blocks sur- vival, adhesion and proliferation of primary chronic lymphocytic leukemia cells. [Published erratum appears in 2016 Leukemia 30: 1963.] Leukemia 30: 337–345.
36. Burger, J. A., and V. Gandhi. 2009. The lymphatic tissue microenvironments in chronic lymphocytic leukemia: in vitro models and the significance of CD40- CD154 interactions. Blood 114: 2560–2561, author reply 2561–2562.
37. Vivier, E., D. H. Raulet, A. Moretta, M. A. Caligiuri, L. Zitvogel, L. L. Lanier,
W. M. Yokoyama, and S. Ugolini. 2011. Innate or adaptive immunity? The example of natural killer cells. Science 331: 44–49.
38. Montresor, A., L. Toffali, A. Rigo, I. Ferrarini, F. Vinante, and C. Laudanna. 2018. CXCR4- and BCR-triggered integrin activation in B-cell chronic lym- phocytic leukemia cells depends on JAK2-activated Bruton’s tyrosine kinase. Oncotarget 9: 35123–35140.
39. Zhu, Z., A. R. Aref, T. J. Cohoon, T. U. Barbie, Y. Imamura, S. Yang,
S. E. Moody, R. R. Shen, A. C. Schinzel, T. C. Thai, et al. 2014. Inhibition of KRAS-driven tumorigenicity by interruption of an autocrine cytokine circuit. Cancer Discov. 4: 452–465.
40. Natarajan, G., S. Oghumu, C. Terrazas, S. Varikuti, J. C. Byrd, and
A. R. Satoskar. 2016. A Tec kinase BTK inhibitor ibrutinib promotes maturation and activation of dendritic cells. OncoImmunology 5: e1151592.
41. Fraietta, J. A., K. A. Beckwith, P. R. Patel, M. Ruella, Z. Zheng, D. M. Barrett,
S. F. Lacey, J. J. Melenhorst, S. E. McGettigan, D. R. Cook, et al. 2016. Ibrutinib enhances chimeric antigen receptor T-cell engraftment and efficacy in leukemia. Blood 127: 1117–1127.
42. Hillmen, P., T. Munir, A. Rawstron, K. Brock, S. Munoz Vicente, F. Yates,
R. Bishop, C. Fegan, D. Macdonald, A. McCaig, et al. 2017. Initial results of ibrutinib plus venetoclax in relapsed, refractory CLL (bloodwise TAP CLARITY study): high rates of overall response, complete remission and MRD eradication after 6 months of combination therapy. Blood 130: 428.
43. Rogers, K. A., Y. Huang, A. S. Ruppert, F. T. Awan, N. A. Heerema, C. Hoffman,
G. Lozanski, K. J. Maddocks, M. E. Moran, M. A. Reid, et al. 2018. Phase 1b study of obinutuzumab, ibrutinib, and venetoclax in relapsed and refractory chronic lymphocytic leukemia. Blood 132: 1568–1572.
44. Lussana, F., M. Cattaneo, A. Rambaldi, and A. Squizzato. 2018. Ruxolitinib- associated infections: a systematic review and meta-analysis. Am. J. Hematol. 93: 339–347.
45. Zeiser, R., and B. R. Blazar. 2017. Pathophysiology of chronic graft-versus-host disease and therapeutic targets. N. Engl. J. Med. 377: 2565–2579.
46. Zeiser, R., A. Burchert, C. Lengerke, M. Verbeek, K. Maas-Bauer,
S. K. Metzelder, S. Spoerl, M. Ditschkowski, M. Ecsedi, K. Sockel, et al. 2015. Ruxolitinib in corticosteroid-refractory graft-versus-host disease after allogeneic stem cell transplantation: a multicenter survey. Leukemia 29: 2062–2068.
47. Miklos, D., C. S. Cutler, M. Arora, E. K. Waller, M. Jagasia, I. Pusic,
M. E. Flowers, A. C. Logan, R. Nakamura, B. R. Momelotinib Blazar, et al. 2017. Ibrutinib for chronic graft-versus-host disease after failure of prior therapy. Blood 130: 2243–2250.