Search

Background:Janus kinase (JAK) inhibitors (JAKinibs) show similar efficacy in rheumatoid arthritis (RA). However, in vitro studies have shown differences in JAK selectivity profiles for baricitinib (BARI), tofacitinib (TOFA), upadacitinib (UPA) and filgotinib (FIL).1,2 These lead to distinct pharmacologic profiles in cellular signaling assays that may impact clinical efficacy or safety1. NK cells are innate lymphocytes important in anti-pathogen responses and immune surveillance, which function via production of cytokines and cell killing3. NK cell proliferation and IFNγ production are JAK-dependent pathways and may be modulated by JAKinibs. Clinical findings show transient decreases in NK cell numbers in patients treated with JAKinibs, but the link to safety is unclear4Objectives:To extend upon findings in proximal cell signaling assays, we compared the selectivity and potency of clinical JAKinibs on NK cell function by assessing proliferation mediated by IL-15 (JAK1/3) and IFN-γ production driven by IL-12 (JAK2/TYK2)+IL-18.Methods:NK cells were isolated from healthy donor PBMC, incubated in vitro with 8 concentrations of each evaluated JAKinib (TOFA, BARI, FIL, FIL metabolite, UPA) and stimulated with IL-15 for proliferation or IL-12/18 for IFNγ production. Proliferation was assessed by Cell Trace dye dilution after 6 days and IFNγ production by intracellular flow cytometry 4hrs post-stimulation. Half maximal inhibitory concentration (IC50) values were calculated for CD56bright, CD56dim, and total NK cells. Steady-state pharmacologic profile over a clinical dosing interval was modeled using concentration-time profiles from JAKinib population pharmacokinetic data in RA subjects under the therapeutic dose5-7. For each JAKinib, the time above IC50 and average daily inhibition of IFNγ or proliferation were calculated for each NK cell population in each donor.Results:Cellular assays in purified NK cells showed dose-dependent inhibition of IL-15-induced proliferation by all JAKinibs with TOFA showing the highest average inhibition and time above IC50 (35-60% inhibition for 8-15 hrs; TOFA>UPA>BARI≈FIL). The differences between JAKinibs are in line with differences in pSTAT inhibition downstream of IL-151. Interestingly, IL-12/18-induced production of IFNγ, which is mediated via JAK2/TYK2 (IL-12) and non-JAK dependent pathways (IL-18), showed weaker inhibition for all compounds. Moreover, all JAKinibs showed 50 for IFNγ production or pSTAT4 inhibition at clinical doses. CD56dim and CD56bright sub-populations of NK cells are proposed to have distinct functions and unique expression of surface receptors. Analysis of the IC50 for pSTAT4 and IFNγ production showed ~2-10-fold weaker inhibition by JAKinibs in CD56bright NK cells, suggesting less dependence on JAK-dependent signals in CD56bright NK cells than CD56dim NK cells.Conclusion:NK cell proliferation depends on JAK1 and JAK3-mediated signaling and is differentially inhibited at clinical doses of distinct JAKinibs. In contrast, functional responses downstream of JAK2/TYK2-dependent IL-12/18 were not substantially inhibited by any of the JAKinibs studied. Inhibition of functional and proliferative responses in purified NK cells aligned well with proximal pSTAT inhibition. JAKinib modulation of NK cell proliferation, but not response to IL-12, reflects unique pharmacologic profiles of the drugs studied and could be one component underlying clinical safety observations, including increased risk of viral infections or malignancy4.References:[1]Traves PG et al. Ann Rheum Dis 2021 (in press)[2]McInnes IB, et al. Arthritis Res Ther 2019;21:183.[3]Cooper MA, Fehniger TA, Caligiuri MA. Trends Immunol 2001 Nov;22(11):633-40.[4]Winthrop KL. Nat Rev Rheumatol 2017; 13(4):234-243[5]Zhang X, et al. CPT Pharmacometrics Syst Pharmacol 2017;6(12):804-13.[6]CDER. Application Number: 203214Orig1s000. NDA 203214: Tofacitinib.[7]Klunder B et al. Clin Pharmacokinet 2019;58(8):1045-58.Disclosure of Interests:Paqui Gonzalez-Traves Shareholder of: Gilead Sciences, Employee of: Gilead Sciences, Laura Simpson Shareholder of: Gilead Sciences, Employee of: Gilead Sciences, Bernard Murray Shareholder of: Gilead Sciences, Employee of: Gilead Sciences, Amy Meng Shareholder of: Gilead Sciences, Employee of: Gilead Sciences, Julie A. Di Paolo Shareholder of: Gilead Sciences, Employee of: Gilead Sciences, Ethan Grant Shareholder of: Gilead Sciences, Employee of: Gilead Sciences, Gundula Min-Oo Shareholder of: Gilead Sciences, Employee of: Gilead Sciences

Background:Janus kinase 1 (JAK1) inhibitors are efficacious in rheumatoid arthritis (RA). Despite having similar efficacy, in vitro studies have shown differences in JAK selectivity profiles for the small-molecule JAK inhibitors (JAKi) baricitinib (BARI), tofacitinib (TOFA), and upadacitinib (UPA).1For example, BARI and UPA are JAK1/JAK2 selective, while TOFA is JAK1/JAK3 selective, but each JAKi has some activity against other JAKs. As JAKs form signaling pairs, differences in selectivity could lead to distinct pharmacologic profiles that may impact clinical efficacy and safety.Objectives:As a first step to understand the basis of potential differences at therapeutic doses, we compared the selectivity and potency of filgotinib (FIL) and its major metabolite (MET) to those of BARI, TOFA, and UPA in cytokine-stimulated peripheral blood mononuclear cells (PBMCs) and whole blood (WB).Methods:PBMCs and WB from healthy donors were incubated in vitro with 8 doses of each JAKi, and levels of signal transducer and activator of transcription phosphorylation (pSTAT) were measured following cytokine stimulation. Half maximal inhibitory concentration (IC50) values were calculated in phenotypically sorted leukocyte populations by flow cytometry. Therapeutic dose relevance of the in vitro analyses was assessed using calculated mean concentration-time profiles from JAKi population pharmacokinetic data in RA subjects. For each JAKi, the time above IC50and average daily pSTAT inhibition were calculated for each cytokine/STAT pair in B cells, CD4+ T cells, CD8+ T cells, monocytes, and/or NK cells.Results:Cellular assays in PBMCs and WB showed dose-dependent inhibition of cytokine-induced pSTATs with all JAKi (correlation between the protein-adjusted IC50values from PBMCs and IC50values from WB, r2=0.98). Among the most potently inhibited pathways were JAK1/TYK2-dependent cytokine, interferon alpha (IFNα), and the JAK1/2-dependent cytokine, interleukin (IL)-6. FIL and MET had weaker potencies against JAK2/TYK2 (G-CSF/pSTAT3), JAK1/2 (IFNƴ/pSTAT1), and JAK2/2 (granulocyte-macrophage colony-stimulating factor [GM-CSF])-dependent pathways compared to JAK1/TYK2 (IFNα/pSTAT5). FIL and MET showed the greatest selectivity vs the JAK2/2 pathway (GM-CSF/pSTAT3) in monocytes.The mean concentration-time profiles and time above IC50over 24 hr for each cytokine/STAT pathway showed that JAK1/2 (IL-6/pSTAT1) and JAK1/TYK2 (IFNα/pSTAT1) pathways were strongly modulated with all tested JAKi. FIL (200 mg) showed similar activity in average target coverage and time above IC50to the approved low doses of TOFA (5 mg) and UPA (15 mg); conversely, FIL had reduced mean average inhibition and time above IC50levels against JAK1/2 (IFNƴ/pSTAT1), JAK1/3-dependent cytokines (IL-2, -4, and -15), JAK2/TYK2 (G-CSF/pSTAT3), and JAK2/2 (GM-CSF/pSTAT5)-dependent pathways compared to TOFA and UPA, and in certain cases to BARI (2 mg).Conclusion:Different JAKi modulate distinct cytokine pathways to varying degrees, and no agent potently and continuously inhibited an individual cytokine signaling pathway throughout the dosing interval. FIL (200 mg) showed a similar inhibition profile to TOFA, BARI, and UPA against the JAK1/TYK2- (IFNα/pSTAT1) or JAK1/2-dependent (IL-6/pSTAT1) responses, consistent with the role of these pathways in clinical efficacy.2However, FIL displayed a differentiated pharmacologic profile from the other JAKi, showing biologically reduced activity on the JAK1/2 (IFNγ)-, JAK1/3 (IL-2, -4 and -15)-, JAK2/TYK2 (G-CSF)-, and JAK2/2 (GM-CSF)-dependent pathways, which play important roles in hematopoiesis and immune function. These data suggest that FIL (200 mg) may have less impact on a subset of homeostatic immune functions signaling via JAK2 and JAK3 than those observed at the clinically approved doses of TOFA (5 mg and 10 mg), UPA (15 mg), and BARI (4 mg).References:[1]McInnes IB, et al. Arthritis Res Ther. 2019;21:183.[2]Banerjee S, et al. Drugs. 2017;77:521-546.Disclosure of Interests:Paqui Gonzalez-Traves Employee of: Gilead, Bernard Murray Employee of: Gilead, Federico Campigotto Employee of: Gilead, Amy Meng Shareholder of: Gilead Sciences, Employee of: Gilead, Julie A. Di Paolo Employee of: Gilead