Targeting GM-CSF in A.B. Avci1, E. Feist2, G.R. Burmester2

1Department of Internal Medicine, ABSTRACT well-known as a haemopoietic growth Rheumatology, Akdeniz University, Granulocyte-macrophage colony-stim- factor used to treat neutropenia follow- Faculty of Medicine, Antalya, Turkey; ulating factor (GM-CSF) is well-known ing . It was a long-time 2Department of Rheumatology and as a haemopoietic growth factor. How- concern that targeted therapies against Clinical Immunology, Charite-University Medicine Berlin, Berlin, Germany. ever, it is also essential in regulating this could cause severe side functions of mature myeloid cells such effects such as neutropenia or pulmo- Ali Berkant Avci, MD, Assoc. Prof. Eugen Feist, PD Dr as macrophages. Preclinical studies nary alveolar proteinosis. Therefore, Gerd-Rüdiger Burmester, Prof. Dr. and observations of flares of arthritis in during the early development phase Please address correspondence to: patients following GM-CSF treatment of compounds targeting GM-CSF or Ali Berkant Avci, MD, supported its important contribution to its receptor special attention was paid Akdeniz Üniversitesi Hastanesi, the pathogenesis of rheumatoid arthritis to this potential adverse event reveal- İç Hastalıkları AD Romatoloji BD, (RA). As the most advanced compound, ing no evidence for such an associated Kampüs 07059 Konyaaltı/Antalya, , a risk profile. On the other hand the so Turkey. against GM-CSF receptor, has already far available results clearly showed E-mail:[email protected] completed phase II trials with a long rapid and sustained effects on disease Received and accepted on June 29, 2016. term of follow-up period of 74 weeks. activity and patient reported outcomes Clin Exp Rheumatol 2016; 34 (Suppl. 98): During this exposure period, an accept- in RA (2). S39-S44. able sustained safety and tolerability © Copyright Clinical and profile has been observed addressing GM-CSF Experimental Rheumatology 2016. the concerns of development of cytope- GM-CSF, a small secreted cytokine is nias or pulmonary alveolar proteinosis. a haemopoietic growth factor responsi- Key words: rheumatoid arthritis, Of note, a rapid and sustained efficacy ble for proliferation and differentiation GM-CSF, mavrilimumab, MOR103, and normalisation of acute phase reac- of myeloid cells from bone marrow tants were consistently shown in studies progenitors. Besides myeloid cells, T both targeting GM-CSF and its recep- and B cells and tissue resident cells in- tor. Its tumour necrosis factor (TNF) cluding chondrocytes, fibroblasts, oste- independent mode of action with con- oblasts, microglia, endothelial and epi- current blockade of GM-CSF as well as thelial cells can secrete GM-CSF (3-5). IL-17 signalling reported from preclini- GM-CSF binds to a heterodimeric GM- cal studies supports the assumption that CSF receptor (GM-CSFR) composed it can be a useful biologic and an alter- of a specific-ligand binding α-chain native agent in TNF inhibitor resistant (GM-CSFRα) and a signal-transducing patients with RA. Therefore, subsequent β-chain (GM-CSFRβ), which is shared studies are warranted to investigate the with IL-3 and IL-5 receptors (6, 7). safety and efficacy of GM-CSF blocking GM-CSFR activation leads to down- agents in different subgroups of RA. stream signalling pathways of -signal transducer and activator Introduction of transcription-suppressor of cytokine Although treatment of rheumatoid ar- signalling (JAK-STAT-SOCS) as well thritis (RA) has evolved tremendously as other pathways including mitogen- in recent years, still an important num- activated protein kinases (MAPK), ber of patients fail to reach the target, phosphatidylinositol 3 kinase (PI3K), disease remission or low disease activ- and NFκB (7-11). The cytoplasmic tail ity (LDA) (1). Blockade of any single of GM-CSFRα has also the capacity to cytokine or cellular subset cannot con- interact with signalling pathways (12). trol the disease in all patients, which GM-CSF is not solely a growth factor necessitates investigations of other cy- responsible for proliferation of mye- Competing interests: G.R. Burmester has tokines or cellular mechanisms respon- loid cells, but also essential in regulat- received honoraria for consulting from sible from RA pathogenesis. To most ing functions of mature myeloid cells MedImmune and GSK; the other authors clinicians, granulocyte-macrophage such as chemotaxis and cell adhesion, have declared no competing interest. colony-stimulating factor (GM-CSF) is dendritic cell function, expression of

S-39 GM-CSF in rheumatoid arthritis / A.B. Avci et al. pro-inflammatory , phago- cytosis, and microbial killing (6, 13, 14). It has a central role in regulating innate immunity. Regarding in vitro studies, GM-CSF can polarise mac- rophages into M1-like macrophages producing inflammatory cytokines in- stead of M2-like macrophages which in contrast produce an anti-inflamma- tory medium (15, 16). However, in vivo studies suggested also association with M2 polarisation in the lung and collaboration of other immune factors such as -gamma (IFNγ), in- terferon regulatory factor 5 (IRF-5) or transforming growth factor-β (TGFβ) family member activin A (17-20). In the pulmonary mucosa, the epithelium is a major producer of GM-CSF. This Fig. 1. GM-CSF as a therapeutic target for inflammation. GM-CSF, is a haemopoietic growth factor re- GM-CSF plays a critical role locally sponsible for proliferation and differentiation of myeloid cells from bone marrow progenitors. Myeloid in regulating microbial defense and cells, T and B cells and tissue resident cells can secrete GM-CSF. GM-CSF can polarise macrophages into M1-like macrophages producing inflammatory cytokines. These cytokines induce the recruitment surfactant clearance by alveolar mac- of inflammatory cells and activation of tissue resident cells. GM-CSF secreted by M1 macrophages rophage population. Defects in GM- induces antigen presenting cells to produce IL-6 and IL-23. IL6 and IL-23 causes again activation of T CSF or its receptor led to death of mice cells and differentiation to TH17 cells which in turn secrete GM-CSF and IL-17 maintaining the circle. GM-CSF produced by TH17cells also induce inflammation by activating monocyte-macrophage system causing pulmonary disease character- and neutrophils. ised by accumulation of surfactant-like APC: antigen presenting cell; DC: dendritic cell; IL: ; G-CSF: granulocyte colony-stimulat- proteins and increased susceptibility to ing factor; GM-CSF: granulocyte-macrophage colony-stimulating factor; TH17: T-helper-17cell; TNF: microbial infection (21). However, a tumour necrosis factor; TRC: tissue resident cell. surprising finding in this study was a relatively normal myelopoiesis. In animal models of arthritis, while de- vrilimumab in adult onset RA patients ficiency and blockade of GM-CSF is with a Disease Activity Score 28-joint GM-CSF and RA protective, overexpression or injection assessment (DAS28) ≤4.8 (mild to GM-CSF is expressed in the synovial of GM-CSF is associated with flares of moderate RA) under a stable dosage membrane and levels of GM-CSF are arthritis (31-35). Other in vitro studies of methotrexate (MTX) for ≥3 months elevated in synovial fluid of RA pa- suggest a role for GM-CSF pathway (43). This was a double-blind placebo- tients (22, 23). The GM-CSFRα is also blockade in tumour necrosis factor controlled study with a 5:1 randomisa- up-regulated in synovial tissue and on (TNF) inhibitor resistant situations and tion (mavrilimumab:placebo) where circulating mononuclear cells from as a part of combinations strategies e.g. patients received single escalating in- RA subjects (24, 25). Synovial tissue with IL-17 or p38 inhibitors (32, 36- travenous (IV) doses of mavrilimumab macrophage populations are associated 39). Finally, the observation of flares (0.01, 0.03, 0.1, 0.3, 1.0, 3.0 and 10.0 with articular damage and decrease in of arthritis following administration of mg/kg) or placebo on study Day 0 and numbers of macrophages is a sensitive recombinant GM-CSF to treat neutro- were followed up for 24 weeks. Over- biomarker for response to treatment in penia in patients with Felty’s syndrome all, 32 patients were enrolled and com- patients with RA (26-28). Since GM- or after chemotherapy suggested a po- pleted the planned evaluations includ- CSF plays a central role in macrophage tential role for targeting GM-CSF and ing safety analysis. The primary objec- differentiation, survival and activation, its receptors in RA (40-42). tive was safety of the compound, while inhibiting GM-CSF activity can affect secondary objectives included pharma- macrophage function and may provide Clinical studies cokinetics and effect on disease activ- clinical benefit in RA. Th17 cells have GM-CSFR inhibition ity. As a surrogate marker for biologic an important place also in RA patho- Phase I. Mavrilimumab (CAM-3001) activity, inhibition of the induction genesis and GM-CSF contributes to is a high-affinity human monoclonal of SOCS3 mRNA by GM-CSF was the differentiation and pathogenicity of antibody to the GM-CSFRα chain shown with doses of 1.0 mg/kg just these cells (29, 30). Figure 1 provides a with competitive antagonistic effect after 4 hours of dosing and this effect schematic representation summarising on GM-CSF signalling. The first in- was sustained over 2 weeks. Of note, the association of GM-CSF with cells human study targeting the innate arm the study population had LDA at base- and cytokines important in RA patho- of the via the GM- line and 63% of patients had normal genesis. CSF pathway was performed with ma- acute phase reactants at the study entry.

S-40 GM-CSF in rheumatoid arthritis / A.B. Avci et al.

Of the n=10 out of 27 subjects treated higher compared to placebo (67.7 vs. formed in adult RA patients who had with mavrilimumab who had elevated 50.7%; p=0.025) with the highest pro- completed EARTH Explorer 1 and 2 erythrocyte sedimentation rate (ESR) portion of moderate or good EULAR Phase IIb studies or were rescued as in- (≥ 20 mm/hr) at baseline, normalised responders in the 100 mg group (46.2% adequate responders at Week 12 (54). values were measured in n=9 of them and 30.8%, respectively). AEs were The aim of the study was to evaluate after treatment. None of the patients generally mild or moderate and no rel- the long-term safety and efficacy of had significant haematological changes evant hypersensitivity reactions, seri- mavrilimumab 100 mg eow through or deterioration of lung functions. Of ous infections or clinically significant 74 weeks of treatment in patients with note, adverse events (AEs) were simi- deterioration of lung functions were moderate to severe RA. Primary end- lar between placebo and mavrilimum- noted. As add-on to the EARTH study, point was to assess the long-term risk ab-treated groups, mostly mild and not the same protocol was performed in benefit ratio of mavrilimumab 100 mg related to the drug dosage. 51 Japanese patients (45). With the eow via evaluation of treatment-emer- Phase II trials. The phase 2A mul- exception of ACR70, the 100mg dose gent adverse events (TEAEs), treat- ticentre, randomised, double-blind, showed a high degree of consistency in ment-emergent serious adverse events placebo-controlled study (EARTH) almost all efficacy endpoints with com- (TESAEs), laboratory parameters and of mavrilimumab included 233 adult parable safety profile to the European pulmonary function tests. Explora- European subjects with moderate-to- patients. tory endpoints included DAS28–CRP severely active seropositive and MTX- A phase IIb study of mavrilimumab LDA (<3.2), DAS28–CRP remission resistant RA (DAS28 ≥3.2) on a stable (EARTH EXPLORER 1) enrolled 326 rate (<2.6), ACR20/50/70 response MTX dose over at least 4 weeks prior patients with moderate to severe adult rates and change in modified Total to screening (44). Patients with active onset RA (DAS28–CRP ≥3.2; ≥4 swol- Sharp Score (mTSS) erosion/narrow- or latent tuberculosis (TB), clinically len joint; mean DAS28-CRP at base- ing scores. Among the 440 patients significant chronic or recurrent infec- line: 5.77) from Europe and South receiving mavrilimumab in this study, tions including hepatitis C or chronic America with an inadequate response the most common TESAEs (n[rate/100 active hepatitis B were excluded. Fur- to at least one of disease-modifying patient years]) were arthritis symptoms thermore, a history of previous treat- anti-rheumatic drugs (DMARDs) (46- (coded as “osteoarthritis”) (4[0,7]) and ment with more than one biologic ther- 52). Patients were randomised to re- bronchitis (3[0,5]). The most common apy for RA that was discontinued for ceive 1 of 3 SC mavrilimumab dosages TEAEs of special interest were bron- lack of efficacy was also an exclusion (150, 100, 30 mg eow) or placebo plus chitis (29[4,8]) and respiratory tract criteria. Subjects received 10, 30, 50 or MTX (7.5–25.0 mg/week). At Week infection(12[2,0]). Within the clinical 100mg of mavrilimumab subcutane- 12, a statistically significant differ- programme, no monocytopenia was re- ous (SC) or placebo every other week ence (p<0.001) in DAS28-CRP as the ported and there were no cases of pul- (eow) for 12 weeks and patients were primary endpoint was seen for all dos- monary alveolar proteinosis. Overall, followed-up for additional 12 weeks. ages of mavrilimumab vs. placebo and also no other pulmonary safety con- The primary endpoint defined as an im- this effect was sustained until Week 24 cerns were identified, and pulmonary provement in DAS28-CRP of at least (p<0.001). Furthermore, a significantly function tests demonstrated a generally 1.2 at Week 12 was significantly more greater percentage of mavrilimumab- transient 20% reduction to <80% of frequently achieved in the mavrili- treated patients met the ACR20 co- predicted values in a few patients. With mumab-treated group compared to pla- primary endpoint vs. placebo for all respect to efficacy, mavrilimumab cebo (55.7% vs. 34.7%; p=0.003). Of dosages at Week 24 (p<0.001). In sum- demonstrated sustained efficacy with note, differences in response rates were mary, mavrilimumab (particularly 150 DAS28-CRP <3.2 and <2.6 rates of observed already by Week 2, increased mg eow) showed rapid and clinically 57.3% and 38.5%, respectively. After throughout the study period and were meaningful effects across a number of 74 weeks of treatment, 68% of patients dose related with best improvement disease activity parameters with an ac- showed no radiographic progression in the 100 mg group. Additionally, ceptable safety and tolerability profile (<0.5 change in mTSS vs. baseline). all categories of the American Col- over the 24-week study period. In conclusion, all performed clinical lege of Rheumatology (ACR) criteria In another phase II study of mavrili- trials reached their primary outcome (ACR20: 69.2% vs. 40.0%, p=0.005; mumab (EARTH EXPLORER 2), and demonstrated the benefit of inhib- ACR50: 30.8% vs. 12.0%, p=0.021; moderate-to-severe active RA patients iting the GM-CSFR-α pathway on RA ACR70: 17.9% vs. 4.0%, p=0.030), as who have had an inadequate response disease activity. Importantly, mavrili- well as the Health Assessment Ques- to one or two anti-TNF agents were mumab showed an acceptable sustained tionnaire Disability Index (HAQ-DI) randomised to SC mavrilimumab safety and tolerability profile, with no (-0.48 vs. -0.25, p=0.005) were sig- 100 mg (eow) or 50 mg significant pulmonary findings over nificantly improved. The proportion of (monthly) for 24 weeks (53). The study the 74-week treatment duration. Since patients achieving a moderate or good has been completed, however, results mavrilimumab produced sustained and European League Against Rheumatism have not been published so far. clinically meaningful effects across a (EULAR) response was significantly An open label extension study was per- number of disease activity parameters

S-41 GM-CSF in rheumatoid arthritis / A.B. Avci et al.

Table I. Clinical trials of agents targeting GM-CSF pathway in rheumatoid arthritis.

Drug/ ClinicalTrials.gov Identifier Target Drug regime Primary outcome measures Trial status

Mavrilimumab NCT00771420 (43) GM-CSFR a single, escalating IV doses of Incidence and severity of AEs Phase I published mavrilimumab (0.01, 0.03, 0.1, 0.3, 1.0, 3.0 and 10.0mg/kg) or placebo, plus MTX Mavrilimumab NCT01050998 (44) GM-CSFR 10mg, 30mg, 50mg, or 100mg SC Improvement in DAS28-CRP Phase II published mavrilimumab eow or placebo, plus MTX Mavrilimumab NCT01050998 GM-CSFR 10mg, 30mg, 50mg, or 100mg SC Improvement in DAS28-CRP Phase II published (Japanese patients) (45) mavrilimumab eow or placebo, plus MTX Mavrilimumab NCT01706926 (46-52) GM-CSFR 150, 100, 30mg SC mavrilimumab eow Improvement in DAS28-CRP Phase IIb abstracts or placebo, plus MTX and ACR20 presented Mavrilimumab NCT01715896 (53) GM-CSFR Mavrilimumab 100mg SC eow or ACR20, 50, 70 responses, Phase II completed, golimumab 50mg SC monthly alternating DAS-28<2,6, HAQ-DI no results posted with placebo, plus MTX improvement>0,25 Mavrilimumab NCT01712399 (54) GM-CSFR Mavrilimumab 100 mg SC eow plus MTX Incidence and severity of AEs Phase II abstract presented MOR103 NCT01023256 (55) GM-CSF Weekly escalating IV doses of MOR103 Incidence and severity of AEs Phase Ib/IIa (0.3, 1.0 and 1.5mg/kg) or placebo published Namilumab NCT01317797 (56) GM-CSF Weekly escalating SC doses of namilumab Incidence and severity of AEs Phase I abstract (150mg and 300mg) or placebo, plus MTX presented Namilumab NCT02354599 (57) GM-CSF Single SC doses of namilumab (80mg, Incidence and severity of AEs Phase I recruiting 150mg and 300mg) or placebo Namilumab NCT02379091 (58) GM-CSF SC doses of namilumab (20mg, 80mg and DAS28-CRP response Phase II recruiting 150mg) or placebo, plus MTX Namilumab NCT02393378 (59) GM-CSF Namilumab 300mg SC at Week 0 followed Change in synovitis, erosion Phase II recruiting by 150mg eow or 40mg eow, and bone marrow edema plus MTX (osteitis) based on OMERACT RAMRIS MORAb-002 NCT01357759 (60) GM-CSF A single, escalating IV doses of Incidence and severity of AEs Phase I completed, MORAb-002 or placebo no results posted KB003 NCT00995449 (61) GM-CSF 70mg, 200mg, or 600mg KB003 IVx5doses Incidence and severity of AEs Phase II terminated or placebo IVx5doses (program refocus)

ACR: American College of Rheumatology; AE: adverse event; CRP: C-reactive protein; DAS28: Disease Activity Score-28; eow: every other week; G-CSF: granulocyte colony-stimulating factor; GM-CSF: granulocyte-macrophage colony-stimulating factor; HAQ-DI: Health Assessment Questionnaire Disability Index; IV: intravenous; MTX: methotrexate; OMERACT: Outcome measures in Rheumatology; RAMRIS: rheumatoid arthritis magnetic resonance imaging studies; SC: subcutaneous.

(DAS28-CRP <3.2, DAS28–CRP <2.6, profile, while secondary endpoints in- in patients with mild to moderate RA ACR20/50/70), further studies are war- cluded DAS28, ACR core set meas- in a phase Ib study, PRIORA (56). Pa- ranted to confirm the promising results ures, EULAR response criteria and tients were on stable MTX doses and in larger cohorts and to find an optimal magnetic resonance imaging of syno- received three single injections of positioning for mavrilimumab in the vitis (MRI synovitis). By using three namilumab 150 or 300mg or placebo treatment algorithm in patients with different dosages of MOR103 (0.3, 1.0 on days 0, 15 and 29 with a follow-up RA. and 1.5 mg/kg) compared to placebo, a period of 12 weeks. Namilumab was rapid and significant clinical improve- generally well tolerated with a similar GM-CSF inhibition ment was demonstrated compared to safety outcome across the treatment MOR103. An alternative approach placebo, most pronounced at the 1.0 groups. By Day 29, greater improve- to interfere with GM-CSF signalling mg/kg dosage in this study. Short-term ments in DAS28 (ESR and CRP) and would be the direct neutralisation of safety and tolerability remained in the joint counts could be observed in nami- the cytokine. The first in patient Study range of placebo. AEs were generally lumab-treated patients compared to of an anti-GM-CSF monoclonal an- mild or moderate and the most com- placebo. Additionally, DAS28-ESR re- tibody (MOR103) in RA included 96 mon AE was nasopharyngitis. sponse rates were higher in namilumab patients with active RA (DAS28≤5.1) Namilumab. Namilumab (MT203) is group than placebo at Day 56 (71.4% (55). The primary outcome measure another monoclonal antibody target- vs. 28.6%). Currently, another phase I was the adverse event rate and safety ing GM-CSF and has been evaluated and two phase II trials of namilumab

S-42 GM-CSF in rheumatoid arthritis / A.B. Avci et al. are recruiting patients (57-59). Current The structure of the GM-CSF receptor com- factor-deficient mice show no major pertur- status and characteristics of the studies plex reveals a distinct mode of cytokine re- bation of hematopoiesis but develop a char- ceptor activation. Cell 2008; 134: 496-507. acteristic pulmonary pathology. Proc Natl targeting GM-CSF pathway are sum- 8. JENKINS BJ, BLAKE TJ, GONDA TJ: Satura- Acad Sci USA 1994; 91: 5592-6. marised in Table I. tion mutagenesis of the beta subunit of the 22. FARAHAT MN, YANNI G, POSTON R, PANAYI human granulocyte-macrophage colony- GS: Cytokine expression in synovial mem- Discussion and conclusion stimulating factor receptor shows clustering branes of patients with rheumatoid arthritis of constitutive mutations, activation of ERK and osteoarthritis. Ann Rheum Dis 1993; 52: Data from both preclinical and clinical MAP kinase and STAT pathways, and differ- 870-5. studies suggest that targeting GM-CSF ential beta subunit tyrosine phosphorylation. 23. BELL AL, MAGILL MK, McKANE WR, KIRK F, and its receptor is a reasonable treat- Blood 1998; 92: 1989-2002. IRVINE AE: Measurement of colony-stimulat- 9. SATO N, SAKAMAKI K, TERADA N, ARAI ing factors in synovial fluid: potential clinical ment option in RA. The favourable K, MIYAJIMA A: Signal transduction by the value. Rheumatol Int 1995; 14: 177-82. safety together with early and sustained high-affinity GM-CSF receptor: two distinct 24. FIELD M, CLINTON L: Expression of GM- efficacy and consistency of results in cytoplasmic regions of the common beta CSF receptor in rheumatoid arthritis. Lancet patients from different continents sup- subunit responsible for different signaling. 1993; 342: 1244. EMBO J 1993; 12: 4181-9. 25. BERENBAUM F, RAJZBAUM G, AMOR B, port the initiation of worldwide phase 10. BROUGHTON SE, NERO TL, DHAGAT U et TOUBERT A: Evidence for GM-CSF recep- III studies. Although the obtained al.: The betac receptor family - Structural tor expression in synovial tissue. An analysis ACR20 and ACR50 responses are com- insights and their functional implications. by semi-quantitative polymerase chain reac- parable to marketed biologic agents, Cytokine 2015; 74: 247-58. tion on rheumatoid arthritis and osteoarthritis 11. HERCUS TR, DHAGAT U, KAN WL et al.: Sig- synovial biopsies. Eur cytokine Netw 1994; 5: the TNF-independent mode of action nalling by the betac family of cytokines. Cy- 43-6. of GM-CSF blockade may provide an tokine Growth Factor Rev 2013; 24: 189-201. 26. MULHERIN D, FITZGERALD O, BRESNIHAN alternative treatment option in patients 12. HERCUS TR, BARRY EF, DOTTORE M, et al.: B: Synovial tissue macrophage populations resistant to other biologics. Moreover, High yield production of a soluble human and articular damage in rheumatoid arthritis. interleukin-3 variant from E. coli with wild- Arthritis Rheum 1996; 39: 115-24. data from preclinical studies support type bioactivity and improved radiolabeling 27. HARINGMAN JJ, GERLAG DM, ZWINDER- the hypothesis that targeting GM-CSF properties. PLoS One 2013; 8: e74376. MAN AH et al.: Synovial tissue macrophages: can be a part of possible future combi- 13. COOK AD, BRAINE EL, HAMILTON JA: Stim- a sensitive biomarker for response to treat- nation therapies especially with concur- ulus-dependent requirement for granulocyte- ment in patients with rheumatoid arthritis. macrophage colony-stimulating factor in in- Ann Rheum Dis 2005; 64: 834-8. rent blockade of IL-17. Finally, the in- flammation. J Immunol 2004; 173: 4643-51. 28. WIJBRANDTS CA, VERGUNST CE, HARING- troduction of new therapies with novel 14. WICKS IP, ROBERTS AW: Targeting GM-CSF MAN JJ, GERLAG DM, SMEETS TJ, TAK PP: targets has the potential to decrease the in inflammatory diseases. Nat Rev Rheuma- Absence of changes in the number of syno- tol 2016; 12: 37-48. vial sublining macrophages after ineffective proportion of refractory patients in RA. 15. FLEETWOOD AJ, LAWRENCE T, HAMILTON treatment for rheumatoid arthritis: Impli- The most important issue in the near JA, COOK AD: Granulocyte-macrophage cations for use of synovial sublining mac- future will be the selection of right bio- colony-stimulating factor (CSF) and mac- rophages as a biomarker. Arthritis Rheum logical DMARD on an individual basis. rophage CSF-dependent macrophage pheno- 2007; 56: 3869-71. types display differences in cytokine profiles 29. CODARRI L, GYULVESZI G, TOSEVSKI V et and transcription factor activities: implica- al.: RORgammat drives production of the References tions for CSF blockade in inflammation. J cytokine GM-CSF in helper T cells, which 1. SMOLEN JS, ALETAHA D: Rheumatoid ar- Immunol 2007; 178: 5245-52. is essential for the effector phase of auto- thritis therapy reappraisal: strategies, oppor- 16. MANTOVANI A, SOZZANI S, LOCATI M, AL- immune neuroinflammation. Nat Immunol tunities and challenges. Nat Rev Rheumatol LAVENA P, SICA A: Macrophage polarization: 2011; 12: 560-7. 2015; 11: 276-89. tumor-associated macrophages as a para- 30. El-BEHI M, CIRIC B, DAI H et al.: The en- 2. AVCI AB, FEIST E, BURMESTER GR: A Promis- digm for polarized M2 mononuclear phago- cephalitogenicity of T(H)17 cells is depend- ing Target in Rheumatoid Arthritis Treatment: cytes. Trends Immunol 2002; 23: 549-55. ent on IL-1- and IL-23-induced production Granulocyte-Macrophage Colony-Stimu- 17. CATES AM, HOLDEN VI, MYERS EM, SMITH of the cytokine GM-CSF. Nat Immunol 2011; lating Factor. Curr Treat Options in Rheum CK, KAPLAN MJ, KAHLENBERG JM: Inter- 12: 568-75. 2015; 1: 320-33. leukin 10 hampers endothelial cell differen- 31. COOK AD, BRAINE EL, CAMPBELL IK, RICH 3. LEIZER T, CEBON J, LAYTON JE, HAMILTON tiation and enhances the effects of interferon MJ, HAMILTON JA: Blockade of collagen- JA: Cytokine regulation of colony-stimu- alpha on lupus endothelial cell progenitors. induced arthritis post-onset by antibody to lating factor production in cultured human Rheumatology 2015; 54: 1114-23. granulocyte-macrophage colony-stimulating synovial fibroblasts: I. Induction of GM-CSF 18. WILLART MA, DESWARTE K, POULIOT P et factor (GM-CSF): requirement for GM-CSF and G-CSF production by interleukin-1 and al.: Interleukin-1alpha controls allergic sen- in the effector phase of disease. Arthritis Res tumor necrosis factor. Blood 1990; 76: 1989- sitization to inhaled house dust mite via the 2001; 3: 293-8. 96. epithelial release of GM-CSF and IL-33. J 32. PLATER-ZYBERK C, JOOSTEN LA, HELSEN 4. CAMPBELL IK, NOVAK U, CEBON J, LAYTON Exp Med 2012; 209: 1505-17. MM, HEPP J, BAEUERLE PA, van den BERG JE, HAMILTON JA: Human articular cartilage 19. KRAUSGRUBER T, BLAZEK K, SMALLIE T WB: GM-CSF neutralisation suppresses in- and chondrocytes produce hemopoietic colo- et al.: IRF5 promotes inflammatory mac- flammation and protects cartilage in acute ny-stimulating factors in culture in response rophage polarization and TH1-TH17 re- streptococcal cell wall arthritis of mice. Ann to IL-1. J Immunol 1991; 147: 1238-46. sponses. Nat Immunol 2011; 12: 231-8. Rheum Dis 2007; 66: 452-7. 5. HAMILTON JA: Colony-stimulating factors in 20. SIERRA-FILARDI E, PUIG-KROGER A, BLAN- 33. CAMPBELL IK, BENDELE A, SMITH DA, inflammation and autoimmunity.Nat Rev Im- CO FJ et al.: Activin A skews macrophage HAMILTON JA: Granulocyte-macrophage munol 2008; 8: 533-44. polarization by promoting a proinflamma- colony stimulating factor exacerbates colla- 6. FLEETWOOD AJ, COOK AD, HAMILTON tory phenotype and inhibiting the acquisition gen induced arthritis in mice. Ann Rheum Dis JA: Functions of granulocyte-macrophage of anti-inflammatory macrophage markers. 1997;56 : 364-8. colony-stimulating factor. Crit Rev Immunol Blood 2011; 117: 5092-101. 34. LANG RA, METCALF D, CUTHBERTSON 2005; 25: 405-28. 21. STANLEY E, LIESCHKE GJ, GRAIL D et al.: RA et al.: Transgenic mice expressing a he- 7. HANSEN G, HERCUS TR, McCLURE BJ et al.: Granulocyte/macrophage colony-stimulating mopoietic growth factor gene (GM-CSF) de-

S-43 GM-CSF in rheumatoid arthritis / A.B. Avci et al.

velop accumulations of macrophages, blind- with rheumatoid arthritis: a randomised, IIb EARTH EXPLORER 1 study [abstract]. ness, and a fatal syndrome of tissue damage. double-blind, placebo-controlled, phase I, EULAR 2015. Ann Rheum Dis 2015; 74 Cell 1987; 51: 675-86. first-in-human study. Ann Rheum Dis 2011; (Suppl. 2): 723. 35. CAMPBELL IK, RICH MJ, BISCHOF RJ, DUNN 70: 1542-9. 53. A study of Mavrilimumab versus anti tumor AR, GRAIL D, HAMILTON JA: Protection from 44. BURMESTER GR, WEINBLATT ME, McINNES necrosis factor in subjects with rheumatoid collagen-induced arthritis in granulocyte- IB et al.: Efficacy and safety of mavrilimum- arthritis. ClinicalTrials.gov [Access date: macrophage colony-stimulating factor-defi- ab in subjects with rheumatoid arthritis. Ann May 15, 2016]; Available from: https://clini- cient mice. J Immunol 1998; 161: 3639-44. Rheum Dis 2013; 72: 1445-52. caltrials.gov/ct2/show/study/NCT01715896 36. HUA F, HENSTOCK PV, TANG B: ERK activa- 45. TAKEUCHI T, TANAKA Y, CLOSE D, GOD- 54. BURMESTER GR, McINNES I, KREMER J et tion by GM-CSF reduces effectiveness of WOOD A, WU CY, SAURIGNY D: Efficacy and al.: Long-term Safety and Efficacy of Ma- p38 inhibitor on inhibiting TNFalpha release. safety of mavrilimumab in Japanese subjects vrilimumab, a Fully Human Granulocyte- Int Immunopharmacol 2010; 10: 730-7. with rheumatoid arthritis: findings from a Macrophage Colony-Stimulating Factor 37. ESPELIN CW, GOLDSIPE A, SORGER PK, Phase IIa study. Mod Rheumatol 2015; 25: Receptor-α (GM–CSFR-α) Monoclonal An- LAUFFENBURGER DA, de GRAAF D, HEN- 21-30. tibody, in Patients with Rheumatoid Arthritis DRIKS BS: Elevated GM-CSF and IL-1beta 46. BURMESTER GR, McINNES IB, KREMER JM (RA) [abstract]. ACR 2015. Arthritis Rheu- levels compromise the ability of p38 MAPK et al.: Efficacy and safety of mavrilimumab, matol 2015; 67 (Suppl. 10). inhibitors to modulate TNFalpha levels in a fully human GM–CSFR-alpha monoclonal 55. BEHRENS F, TAK PP, OSTERGAARD M et al.: the human monocytic/macrophage U937 cell antibody in patients with rheumatoid arthri- MOR103, a human monoclonal antibody to line. Mol Biosyst 2010; 6: 1956-72. tis: primary results from the EARTH EX- granulocyte-macrophage colony-stimulating 38. van NIEUWENHUIJZE AE, van de LOO FA, PLORER 1 study [abstract]. EULAR 2015: factor, in the treatment of patients with mod- WALGREEN B et al.: Complementary action Ann Rheum Dis 2015; 74 (Suppl. 2): 78. erate rheumatoid arthritis: results of a phase of granulocyte-macrophage colony-stimu- 47. KREMER JM, BURMESTER GR, WEINBLATT Ib/IIa randomised, double-blind, placebo- lating factor and interleukin-17A induces M et al.: Patient-reported outcomes (PROS) controlled, dose-escalation trial. Ann Rheum interleukin-23, receptor activator of nuclear during treatment with mavrilimumab, a fully Dis 2015; 74: 1058-64. factor-kappaB ligand and matrix metallo- human monoclonal antibody targeting GM– 56. HUIZINGA TWJ, BATALOV A, YABLANSKI K proteinases and drives bone and cartilage CSFR-alpha, in the phase IIb EARTH EX- et al.: First-in-patient study of namilumab, an pathology in experimental arthritis: rationale PLORER 1 study [abstract]. EULAR 2015. anti-GM-CSF monoclonal antibody, in active for combination therapy in rheumatoid ar- Ann Rheum Dis 2015; 74 (Suppl. 2): 483. rheumatoid arthritis: results of the PRIORA thritis. Arthritis Res Ther 2015; 17: 163. 48. McINNES IB, BURMESTER GR, KREMER JM phase Ib study [abstract]. EULAR 2015. Ann 39. PLATER-ZYBERK C, JOOSTEN LA, HELSEN et al.: Rapid Onset of Clinical Benefit Is As- Rheum Dis 2015; 74 (Suppl. 2): 733. MM, KOENDERS MI, BAEUERLE PA, van den sociated with a Reduction in Validated Bio- 57. A Phase 1 MT203 Single-dose Study to BERG WB: Combined blockade of granulo- markers of Disease in Patients with Rheuma- Evaluate Safety, PK and PD. ClinicalTrials. cyte-macrophage colony stimulating factor toid Arthritis Treated with Mavrilimumab, gov [Access date: May 15, 2016]; Available and interleukin 17 pathways potently sup- a Human Monoclonal Antibody Targeting from: https://clinicaltrials.gov/ct2/show/ presses chronic destructive arthritis in a tu- GM-CSFR [abstract]. ACR 2014. NCT02354599. mour necrosis factor alpha-independent mouse 49. GUO X, SINIBALDI D, KUZIORA M et al.: 58. Dose Finding Study of Namilumab in Com- model. Ann Rheum Dis 2009; 68: 721-8. Sustained response to mavrilimumab in rheu- bination With Methotrexate in Participants 40. PEREIRA J, VELLOSO ED, LOTERIO HA, LAU- matoid arthritis patients via suppression of With Moderate to Severe Rheumatoid Ar- RINDO IM, CHAMONE DA: Long-term remis- macrophage and T cells [abstract]. EULAR thritis (RA). ClinicalTrials.gov [Access date: sion of neutropenia in Felty’s syndrome after 2015. Ann Rheum Dis 2015; 74 (Suppl. 2): May 15, 2016]; Available from: https://clini- a short GM-CSF treatment. Acta Haematol 734. caltrials.gov/ct2/show/NCT02379091. 1994; 92: 154-6. 50. JIN DC, WU CY, ROSKOS LK, GODWOOD 59. Namilumab vs Adalimumab in Participants 41. HAZENBERG BP, Van LEEUWEN MA, Van RI- A, CLOSE D, WANG B: Exposure–efficacy With Moderate to Severe Early Rheuma- JSWIJK MH, STERN AC, VELLENGA E: Cor- analysis of mavrilimumab in rheumatoid ar- toid Arthritis Inadequately Responding to rection of granulocytopenia in Felty’s syn- thritis: modeling and simulation of phase II Methotrexate (TELLUS). ClinicalTrials. drome by granulocyte-macrophage colony- clinical data [abstract]. EULAR 2015. Ann gov [Access date: May 15, 2016]; Available stimulating factor. Simultaneous induction Rheum Dis 2015; 74 (Suppl. 2): 1043. from: https://clinicaltrials.gov/ct2/show/ of interleukin-6 release and flare-up of the 51. KREMER JM, BURMESTER GR, WEINBLATT NCT02393378. arthritis. Blood 1989; 74: 2769-70. M et al.: Analysis of Patient-Reported Out- 60. Safety and Tolerability of MORAb-022 in 42. de VRIES EG, WILLEMSE PH, BIESMA B, comes during Treatment with Mavrilimumab, Healthy and Rheumatoid Arthritis Subjects. STERN AC, LIMBURG PC, VELLENGA E: a Human Monoclonal Antibody Targeting ClinicalTrials.gov [Access date: May 15, Flare-up of rheumatoid arthritis during GM- GM-CSFRá, in the Randomized Phase 2b 2016]; Available from: https://clinicaltrials. CSF treatment after chemotherapy. Lancet Earth Explorer 1 Study [abstract]. ACR 2014. gov/ct2/results?term=NCT01357759. 1991; 338: 517-8. 52. McINNES IB, BURMESTER GR, KREMER JM 61. Study of KB003 In Biologics-Inadequate 43. BURMESTER GR, FEIST E, SLEEMAN MA, et al.: Rapid onset of clinical benefit in pa- Rheumatoid Arthritis. ClinicalTrials. WANG B, WHITE B, MAGRINI F: Mavrili- tients with RA treated with mavrilimumab, a gov [Access date: May 15, 2016]; Avail- mumab, a human monoclonal antibody tar- fully human monoclonal antibody targeting able from: https://clinicaltrials.gov/ct2/ geting GM-CSF receptor-alpha, in subjects GM–CSFR-alpha: subanalysis of the phase results?term=NCT00995449.

S-44