P216 Effects of filgotinib on anaemia, thrombocytopoenia and leukopoenia: results from a Phase 3 study in patients with active RA and prior inadequate response or intolerance to bDMARDs

Background Cytopoenias are common in patients treated for rheumatoid arthritis (RA) with non-janus kinase 1 (JAK1)-selective inhibitors, possibly due to JAK2-mediated haematopoietic growth factor inhibition. We investigated the extent of cytopoenia in patients with active RA, despite prior treatment with biological disease-modifying antirheumatic drugs (bDMARDs), treated with the JAK1-selective inhibitor filgotinib (FIL), in a Phase 3 trial (FINCH2; NCT02873936). Methods In the double-blind, Phase 3 FINCH2 trial, patients were randomised 1:1:1 to receive oral FIL 200mg, 100mg, or placebo (PBO) once daily for 24 weeks (W) + conventional synthetic DMARDs. We assessed shifts from baseline at 12 and 24 weeks in haemoglobin, platelets, neutrophils and lymphocytes. Results 448 patients were treated: FIL 200mg, n = 147; FIL 100mg, n = 153; PBO, n = 148. Overall, haemoglobin, platelet, lymphocyte and neutrophil levels remained consistent throughout the study. At baseline, 129 (28.8%), 4 (0.9%), 10 (2.2%) and 26 (5.8%) patients had mild-moderate low levels of haemoglobin, platelets, neutrophils and lymphocytes, respectively, and 5 (1.1%) had severely low levels of lymphocytes. Of the patients with mild-moderate low haemoglobin levels at baseline, 10-13% achieved normal levels by W24 vs 8% receiving PBO (Table). Of those with normal baseline haemoglobin levels, 6-10% had mild low levels at W24. All patients with baseline mild-moderate low platelets and neutrophils had normal levels at W24, except one patient with mild neutropoenia receiving FIL 100mg. Of the patients with normal platelet and neutrophil levels at baseline, >94% maintained these at W24 in all treatment groups. By W24, 3.2%, 5.2% and 2.2% of patients treated with FIL 200mg, FIL 100mg and PBO, respectively in the baseline mild-moderate subgroup and 1.7% in the severe subgroup treated with FIL 100mg had normal lymphocyte counts. Conclusion In this study, most patients in the baseline normal cell count subgroups maintained this status over 24 weeks of FIL treatment. Of the patients with mild-to-moderately low haemoglobin at baseline, >9% shifted towards haemoglobin normalisation. Similar patterns of improvement from baseline were observed for platelet, lymphocyte and neutrophil counts. FIL appears not to increase the incidence of cytopenias in patients with active RA despite prior biologic therapies. Disclosures D. Walker: Other; Received support from Lilly, Pfizer, Novartis, Roche. M.C. Genovese: Other; Received support from Gilead Sciences Inc., Galapagos NV, AbbVie Inc. Eli Lilly and Company, Pfizer. K. Kalunian: Grants/research support; Grand support from Gilead. J. Gottenberg: None. B. Bartok: Corporate appointments; Employee of Gilead Sciences, Inc. Shareholder/stock ownership; Shareholder of Gilead Sciences, Inc. Y. Tan: Corporate appointments; Employee of Gilead Sciences, Inc... Shareholder/stock ownership; Shareholder of Gilead Sciences, Inc. Y. Guo: Corporate appointments; Employee of Gilead Sciences, Inc. Shareholder/stock ownership; Shareholder of Gilead Sciences, Inc. C. Tasset: Other; Employee of Galapagos. J.S. Sundy: Corporate appointments; Employee of Gilead Sciences, Inc. Shareholder/stock ownership; Shareholder of Gilead Sciences, Inc. K. de Vlam: None. T. Takeuchi: None.

Abstract 549: Extracellular Heat Shock Protein 27 Signals Through NF-κB to Favorably Modulate Macrophage Inflammation

Background: Previously we demonstrated that recombinant Heat Shock Protein 27 (rHSP27) reduces atherosclerotic lesion formation when administered to ApoE -/- mice. In addition, administration of rHSP27 to ApoE -/- mice with established atherosclerotic lesions halts lesion progression and is associated with lesion modifications that are consistent with resilience to plaque rupture. However, the mechanism(s) for these therapeutic effects remain elusive. Objective: To determine whether rHSP27 favorably modulates macrophage inflammation by focusing on NF-kB signaling. Methods/Results: Activation of the NF-kB signaling pathway by rHSP27 was observed in peritoneal macrophages from ApoE -/- mice. Treatment with rHSP27 for 30 minutes activated the translocation of the NF-kB p65 subunit from the cytosol to the nucleus as observed by immunolabeling. A dose-dependant increase in rHSP27 mediated NF-kB activation was observed in RAW 264.7 macrophages stably transfected with an NF-kB inducible reporter gene (15 fold; p<0.05). The use of an N-terminal deletion mutant of rHSP27, rC1, at equimolar concentrations did not induce NF-kB activation, demonstrating specificity of the full-length protein. In addition, NF-kB inhibitors (BAY 11-7082 and MG-132) attenuated the induction of the NF-kB reporter gene by rHSP27, thereby implicating the involvement of IkBa phosphorylation and degradation by the proteasome. A consequence of rHSP27 signaling in macrophages was the up-regulation of the transcript for the haematopoietic growth factor/regulator, GM-CSF (300 fold; p<0.05) and subsequently its secretion (400 fold, p<0.05). In the presence of BAY, GM-CSF expression was inhibited suggesting the involvement of NF-kB. Conclusions: rHSP27 promotes the nuclear translocation and activation of NF-kB in macrophages which results in the up-regulation of GM-CSF, a haematopoietic growth factor/regulator that may be responsible for the observed therapeutic effects of rHSP27 in vivo. Current studies are testing rHSP27 therapy in ApoE -/- mice deficient in GM-CSF in order to evaluate the importance of macrophage modulation for the beneficial affects of rHSP27 in atherogenesis.

TPO Exerts a Protective Effect on Iron-Induced Apoptosis Via Erk1/2 Signaling in Human Mesenchymal Stem Cells.

Osteoporosis is a very common problem among adolescent and adult patients with thalassaemia major (TM). The pathogenesis of osteoporosis in TM is related to several factors including iron overload. Bone is derived from osteoblasts. And osteoblasts are differentiated from mesenchymal stem cells (MSC). Therefore, iron-overload induced MSC damage may contribute to osteopenia and osteoporosis. The effect of iron-overload on MSC has not been investigated previously. We hypothesize that iron-overload may induce apoptosis in MSC by caspase-dependent pathway, and haematopoietic growth factor thrombopoietin (TPO) and calcium channel blocker amlodipine may have a protective effect on iron-induced apoptosis in these cells. We have shown that iron (FeCl3) reduced hMSCs viability in a dose-dependent manner (0–0.6 mM) (n=5). By annexin V and PI staining, apoptotic cells were found to be significantly increased after iron treatment (0.3 mM) for 72 hrs (n=4). The expression of active caspase-3 was significantly increased in iron-treated cells (0.15mM, 0.3mM) (n=5). Iron treatment also increased the proportion of cells containing JC-1 monomers, indicating a trend in the drop of mitochondrial membrane potential. TPO exerted protective effect on iron-induced apoptosis in hMSCs. Human MSCs were cultured in the presence of iron (0.3 mM) with or without TPO (50 ng/ml) for 72 hrs (n=4). The cell viability was significantly increased with the treatment of TPO. Dot-plot analysis of annexin V/PI staining demonstrated that TPO significantly reduced the population of apoptotic cells. Incubation with TPO also decreased the iron-induced caspase-3 expression. Flow cytometric dot-plot analysis of hMSCs also showed trends of amelioration of the increase in JC-1 monomers in the iron plus TPO. The population of phospho-Erk1/2 was also significantly increased in TPO-treatment, and the increased phospho-Erk was significantly reversed by the upstream signaling inhibitor PD098059. Calcium channel blocker amlodipine (10−9M) also had a protective effect on iron-induced apoptosis in these cells. Our findings suggest that iron-overload induces apoptosis in hMSCs via the caspase-dependent pathway and that TPO and amlodipine might exert a protective effect on iron-induced apoptosis via the activation of Erk1/2 signaling. The use of either haematopoietic growth factor or calcium channel blocker for the protection of hMSCs from iron induced toxicity is a novel concept. Our study has the potential in minimizing the bone damage induced by iron-overload in patients with thalassaemia major.

Human thrombopoietin structure–function relationships: identification of functionally important residues

Thrombopoietin (TPO) is a haematopoietic growth factor responsible for megakaryocyte progenitor proliferation and differentiation. It belongs to the four-helix-bundle cytokine family and exerts its biological effects through binding to a specific receptor, c-Mpl. With the use of site-directed mutagenesis we have generated 20 TPO mutants. Each of the TPO mutants was produced in a eukaryotic expression system and the mutants' ability to induce the proliferation of factor-dependent c-Mpl-expressing megakaryoblastic M-O7e cells was compared with that of wild-type TPO. Among the mutations studied, 10 lead to a significant decrease in TPO bioactivity. Of these ten residues, three are located in helix A of the protein (Arg10, Lys14 and Arg17) and four in helix D (His133, Gln132, Lys138 and Phe141), indicating that in TPO, as in other cytokines, these two helices are important for functional cytokine/receptor interactions. Surprisingly, mutant Arg10 → Ala (R10A) lacked any proliferative activity, despite the fact that this mutation was recently reported to have no effect on TPO/c-Mpl binding in a TPO phage ELISA [Pearce, Potts, Presta, Bald, Fendly and Wells (1997) J. Biol. Chem. 272, 20595–20602]. The lack of M-O7e proliferation is probably due to an inability of R10A mutant to promote receptor dimerization and thus receptor activation. Moreover we found that the Arg10 and Arg17 residues of TPO seem to be specific determinants for TPO/c-Mpl recognition. We also demonstrate that the O-glycosylation site located at position 110 of TPO is not necessary for the bioactivity of the cytokine.