Efficacy/Effectiveness, Safety and Treatment Persistence of Filgotinib in Rheumatoid Arthritis: A Narrative Review of Clinical Trials and Real-World Evidence

Abstract

Filgotinib (FIL) is a Janus kinase preferential inhibitor, classified as a targeted synthetic disease-modifying anti-rheumatic drug (tsDMARD). It is approved for patients with rheumatoid arthritis (RA) who have failed treatment with methotrexate (MTX), a conventional synthetic DMARD (csDMARD). Randomized controlled trials (RCTs) investigating 100 and 200 mg doses of FIL + MTX have demonstrated rapid effects in reducing disease activity, pain, and fatigue and improving functional outcomes compared with MTX monotherapy or adalimumab (a biologic) + MTX. FIL’s efficacy in reducing disease activity and improving functional outcomes was sustained in the long-term (3 years), while also minimizing radiographic progression. In particular, 200 mg FIL + MTX significantly reduced radiographic progression uncoupled from disease activity, outperforming adalimumab + MTX in patients with medium or high disease activity. Adverse events were similar among FIL, MTX, and adalimumab over 52 weeks in RCTs, with a consistent safety profile in the long term (> 8 years). FIL persistence rates were generally high, regardless of dose. Evidence from real-world observational studies complements and underscores the effectiveness and safety of FIL demonstrated in RCTs. These findings support FIL’s potential as a therapeutic option in the management of RA. Guidelines currently recommend a starting dose of 200 mg FIL ± MTX; a 100 mg starting dose can be used if the patient is elderly (Europe); at increased risk of venous thromboembolism, major adverse cardiovascular events or malignancy (Europe); or has impaired renal function (Japan). Furthermore, patients should only initiate FIL after inadequate response to MTX or other csDMARDs. In clinical practice, there may be additional considerations that could impact outcomes, including persistence and variations in patient profiles and treatment backgrounds. Future results from ongoing studies can be expected to provide further insights into the long-term effectiveness, such as joint outcomes and safety of FIL, in the real-world context.

Graphical abstract available for this article.

Plain Language Summary

Filgotinib is an oral medication for rheumatoid arthritis. It is a Janus kinase inhibitor that blocks a protein involved in long-term inflammation. Patients with rheumatoid arthritis are usually first treated with methotrexate, a conventional medication. Filgotinib is for patients whose methotrexate or other similar treatments did not work well enough, or for those who cannot tolerate these treatments. Clinical trials have shown that filgotinib can provide better, faster effects, including pain relief, than methotrexate alone. Additionally, filgotinib helps prevent joint damage and improves movement. Most people can take filgotinib without serious side effects; in clinical trials, filgotinib had similar side effects as methotrexate. Therefore, patients could continue filgotinib without stopping treatment early. While clinical trials are conducted in more controlled environments, studies of filgotinib in real-world settings showed similar effectiveness and safety. Currently, filgotinib is approved for treatment of rheumatoid arthritis in the European Union, UK, Japan, Taiwan, South Korea and Singapore. In Europe and Japan, the recommended starting dose of filgotinib is 200 mg, either alone or with methotrexate. A lower starting dose may be needed for certain patients; 100 mg could be advised for the elderly or those at higher risk of blood clots, heart issues or cancer, or with kidney problems. Other factors that may affect how well filgotinib works include prior treatments and other conditions patients may have besides rheumatoid arthritis. Further research will shed more light on real-life results with filgotinib in the future.

Graphical Abstract

Key Summary Points

Findings of randomized controlled trials (RCTs) have demonstrated the rapid and sustained efficacy (clinical, structural and functional endpoints) and safety of filgotinib in rheumatoid arthritis

Evidence from real-world studies complements and underscores the effectiveness and safety of filgotinib demonstrated in RCTs, including high rates of treatment persistence

Japanese and European guidelines recommend a starting dose of 200 mg filgotinib with or without methotrexate following initial treatment with methotrexate or other conventional synthetic disease-modifying anti-rheumatic drugs

A dose of 100 mg filgotinib is recommended in patients who are elderly, or at increased risk of venous thromboembolism, major adverse cardiovascular events or malignancy (Europe) or have impaired renal function (Japan)

The above comorbidities are a key consideration in screening patients for appropriate management of rheumatoid arthritis using filgotinib, especially older patients

Digital Features

This article is published with digital features, including a graphical abstract to facilitate understanding of the article. To view digital features for this article, go to 10.6084/m9.figshare.31231801.

Introduction

Rheumatoid arthritis (RA) is an autoimmune disease characterized by symmetrical, invasive inflammatory lesions in multiple joints. It can severely damage and deform joints and may lead to progressive disability [1]. The global burden of RA continues to grow; in 2021, an estimated 17.9 million people were living with RA, representing a 125% increase compared to 1990 [1]. In 2021, Japan (~ 475,000 cases) and the UK (UK; ~ 275,000 cases) were among the countries with the highest numbers of prevalent RA cases [2]; within the UK, the incidence of RA has increased by 1.64 times in the 17 years between 2002 and 2019 [3].

The widespread implementation of a treat-to-target approach to pharmacological management and an expanded repertoire of highly effective disease-modifying anti-rheumatic drugs (DMARDs) have transformed RA treatment by suppressing autoimmune activity to delay or prevent joint degradation [4]. Conventional synthetic DMARDs (csDMARDs) such as methotrexate (MTX) are typically the first course of treatment for newly diagnosed patients and are currently recommended for initial treatment by the 2021 American College of Rheumatology (ACR), 2022 European Alliance of Associations for Rheumatology (EULAR) and 2022 Japan College of Rheumatology (JCR) guidelines [5–7].

Over the past 2 decades, biologic DMARDs (bDMARDs) and targeted synthetic DMARDs (tsDMARDs) have been introduced, providing more specific targeting of inflammatory pathways than broad-acting csDMARDs [4]. bDMARDs are genetically engineered monoclonal antibodies or recombinant fusion proteins, further divided into separate classes depending on their mechanism of action, such as tumor necrosis factor-alpha (TNF-α) inhibitors and interleukin-6 receptor (IL-6R) inhibitors [4, 8]. tsDMARDs are inhibitors of specific inflammatory signaling pathways and intracellular processes [9]. To date, the only tsDMARDs approved for RA treatment are those targeting Janus kinase (JAK) cytoplasmic proteins to modulate the inflammatory response [9]. These JAK inhibitors (JAKis) have the unique advantage of oral administration, an attribute strongly valued by many patients [10], compared with bDMARDs, which are intravenously or subcutaneously administered [4]. bDMARDs and tsDMARDs are recommended in the 2022 EULAR and 2024 JCR guidelines for patients with poor prognostic factors (PPFs) or if patients have an inadequate response or are intolerant to MTX [6, 11].

Filgotinib (FIL) is a JAKi with high selectivity for JAK1 over other JAK isoforms (JAK2, JAK3) [4]; once absorbed and metabolized, FIL is also biotransformed into an active metabolite that is selective for JAK1, though with lower affinity [12]. FIL is included in treatment recommendations as an option for patients with inadequate response to csDMARDs or bDMARDs, preferably in combination with MTX [6, 11]. These recommendations were based on evidence from randomized controlled trials (RCTs), such as FINCH, which have demonstrated that FIL is a well-tolerated and efficacious option for patients with moderate-to-severe RA, including inadequate responders (IRs) to first-line MTX [6, 13]. The JCR FIL-specific guidelines recommend a daily dose of 200 mg FIL, with a lower 100 mg FIL dose being considered for patients with a lower estimated glomerular filtration rate (eGFR), ≤ 60 ml/min/1.73 m² [11, 14].

In addition to RCT data, observational studies on FIL use among patients with RA in the real-world setting have been conducted. This narrative review aims to summarize the safety, efficacy and real-world effectiveness of FIL for the treatment of RA based on a discussion of findings from clinical trials and observational studies. As this review is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors, ethical approval is not required.

Efficacy and Effectiveness of Filgotinib Regarding Clinical Outcomes

Rapid Effectiveness

Clinical effectiveness of RA treatment is usually assessed in terms of control of disease activity, using measures such as disease activity indices or clinical response rates. As uncontrolled inflammation is a strong predictor of disease progression and functional impairment [15], rapid effectiveness (e.g., in the first 6 months) of treatment is a key consideration in the management of RA.

A blood biomarker analysis of FINCH 1 has demonstrated a significant positive correlation between disease activity and cytokines associated with RA pathobiology [16]. FIL + MTX reduced levels of cytokines involved in inflammation and cell migration (IL-6, CXCL13 and TNF-α) as early as Week 4 and sustained this reduction through to Week 12 in a dose-dependent manner [16]. While significant reductions in these cytokines were also observed in patients receiving adalimumab (a bDMARD), patients receiving placebo + MTX (i.e., MTX monotherapy) had similar levels of these cytokines between baseline and Weeks 4 and 12 [16]. Additionally, significant reductions in bone-related turnover biomarkers, such as N-telopeptide of type 1 collagen and C-telopeptide 1, were observed at Week 12 with FIL compared with adalimumab, which had less pronounced effects on these biomarkers [16].

The rapid effectiveness of FIL, as shown in the biomarker analysis, is reflected in the early reductions in disease activity observed in the FINCH RCTs [17–20], with significantly higher proportions of patients with RA on FIL + MTX achieving remission by Week 12 compared with MTX monotherapy. Among IRs to MTX from FINCH 1, proportions of patients achieving remission (defined using Disease Activity score including 28 joints using C-reactive protein [DAS28CRP], Clinical Disease Activity Index [CDAI] and Simplified Disease Activity Index [SDAI]) at Week 12 were significantly greater for FIL (100 mg or 200 mg) + MTX compared with MTX monotherapy (Table 1) [18, 20]. In a subgroup analysis of patients with 4 PPFs (e.g., seropositivity for rheumatoid factors or anti-cyclic citrullinated peptide, serum high-sensitivity C-reactive protein > 6 mg/l, DAS28CRP > 5.1 and erosions), FIL (100 mg or 200 mg) + MTX also had significantly higher DAS28CRP remission rates as well as numerically higher SDAI and CDAI remission rates at Week 12 (Table 1) [18]. Further post hoc analyses of FINCH 1 (IRs to MTX) and FINCH 3 (patients who were MTX-naïve) showed that FIL (100 mg or 200 mg) + MTX achieved remission at Week 24 (defined using DAS28CRP, CDAI or SDAI) regardless of estimated rapid radiographic progression (eRRP; defined as ≥ 5 baseline modified total Sharp score [mTSS] or based on duration of RA; Table 1) [17].

Table 1.

Summary of rapid efficacy and effectiveness findings for FIL from RCTs and real-world studies

Effectiveness measure	Timepoint	Population	Result
DAS28CRP reduction (mean ± SD score or median [IQR])	Month 1	FILOSOPHY; European prospective observational study: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 4.2 (2.7–5.5)
	Week 12	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naive: • 200 mg FIL ± MTX: 2.8 ± 1.3 (p < 0.05 from baseline) • 200 mg FIL: 2.9 ± 1.4 (p < 0.05 from baseline) • 200 mg FIL + MTX: 2.6 ± 1.2 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 3 ± 1.6 (p-value NR) • 200 mg FIL: 3.1 ± 1.8 (p-value NR) • 200 mg FIL + MTX: 3 ± 1.3 (p-value NR)
	Week 24	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naive: • 200 mg FIL ± MTX: 2.4 ± 1.4 • 200 mg FIL: 3 ± 2.5 • 200 mg FIL + MTX: 2 ± 0.9	bDMARD-IR: • 200 mg FIL ± MTX: 2.6 ± 1.5 • 200 mg FIL: 2.4 ± 1.6 • 200 mg FIL + MTX: 2.7 ± 1.4
DAS28CRP remission (proportion of patients)	Week 12	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Moderate RA: • MTX monotherapy: 15.6% (reference group) • 200 mg FIL + MTX: 47.1% (p < 0.0001) • 100 mg FIL + MTX: 37.2% (p < 0.0001) • Adalimumab + MTX: 44.4% (p value NR)
			Moderate-to-severe RA: • MTX monotherapy: 9.3% (44/475) (reference group) • 200 mg FIL + MTX: 34.1% (n = 162/475) (p < 0.001) • 100 mg FIL + MTX: 23.8% (n = 114/480) (p < 0.001) • Adalimumab + MTX: 23.7% (n = 77/325) (p < 0.001)
			4 PPFs: • MTX monotherapy: 6.1% (n = 11/181) (reference group) • 200 mg FIL + MTX: 30.9% (n = 59/191) (p < 0.001) • 100 mg FIL + MTX: 19.6% (n = 37/189) (p < 0.001) • Adalimumab + MTX: 15.9% (n = 20/126) (p-value NR)	< 4 PPFs: • MTX monotherapy: 11.2% (n = 33/294) (reference group) • 200 mg FIL + MTX: 36.3% (n = 103/284) (p < 0.001) • 100 mg FIL + MTX: 26.5% (n = 77/291) (p < 0.001) • Adalimumab + MTX: 28.6% (n = 57/199) (p-value NR)
	Week 24	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 55.3% vs 55.5% (p = 1.00) • 100 mg FIL + MTX: 36.5% vs 42.2% (p = 0.29) • Adalimumab + MTX: 32.6% vs 45.5% (p = 0.05) • MTX monotherapy: 17.6% vs 22.2% (p = 0.39)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 59.3% vs 61.5% (p = 0.74) • 100 mg FIL + MTX: 45.9% vs 47.1% (p = 0.88) • Adalimumab + MTX: 51.3% vs 43.8% (p = 0.36) • MTX monotherapy: 34.5% vs 30.3% (p = 0.43)
	Month 3	Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 46.3% (n = 38/82) (p < 0.0001 from baseline)
			tsDMARD-exposed: • 200 mg FIL ± MTX: 36.4%
			bDMARD-exposed: • 200 mg FIL ± MTX: 48%	bDMARD-naïve: • 200 mg FIL ± MTX: 36.8%
		Schultz et al.; German multicenter retrospective observational study: • Moderate-to-severe RA	Overall: • 200 mg FIL ± MTX: n = 84/216
	Month 6	Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 66.2% (n = 45/82) (p < 0.0001 from baseline, p = 0.021 from Month 3)
			tsDMARD-exposed: • 200 mg FIL ± MTX: 50%
			bDMARD-exposed: • 200 mg FIL ± MTX: 64.3%	bDMARD-naïve: • 200 mg FIL ± MTX: 69.2%
		Schultz et al.; German multicenter retrospective observational study: • Moderate-to-severe RA	Overall: • 200 mg FIL ± MTX: n = 110/167
DAS28ESR remission (proportion of patients)	Month 3	Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall • 200 mg FIL ± csDMARDs: 29% (p < 0.001 from baseline)
CDAI reduction (mean ± SD score)	Week 12	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naive: • 200 mg FIL ± MTX: 9.3 ± 4.9 (p < 0.05 from baseline) • 200 mg FIL: 9.8 ± 6 (p-value NR) • 200 mg FIL + MTX: 8.9 ± 4.1 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 13.3 ± 8.3 (p-value NR) • 200 mg FIL: 15 ± 9.3 (p-value NR) • 200 mg FIL + MTX: 11.8 ± 6.9 (p-value NR)
	Week 24	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naive: • 200 mg FIL ± MTX: 7.5 ± 7.5 (p < 0.05 from baseline) • 200 mg FIL: 10.2 ± 2.9 (p-value NR) • 200 mg FIL + MTX: 5.9 ± 1.9 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 11.3 ± 7.8 (p-value NR) • 200 mg FIL: 10.9 ± 7.9 (p-value NR) • 200 mg FIL + MTX: 11.5 ± 7.8 (p-value NR)
CDAI remission (proportion of patients)	Week 12	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Moderate RA: • MTX monotherapy: 3.9% (reference group) • 200 mg FIL + MTX: 14.4% (p = 0.010) • 100 mg FIL + MTX: 16.5% (p = 0.003) • Adalimumab + MTX: 9.7% (p-value NR)
			4 PPFs: • MTX monotherapy: 2.2% (n = 4/181) (reference group) • 200 mg FIL + MTX: 9.4% (n = 18/191) (p = 0.004) • 100 mg FIL + MTX: 9.5% (n = 18/189) (p = 0.003) • Adalimumab + MTX: 4.8% (n = 6/126) (p-value NR)	< 4 PPFs: • MTX monotherapy: 3.1% (n = 9/294) (reference group) • 200 mg FIL + MTX: 14.4% (n = 41/284) (p < 0.001) • 100 mg FIL + MTX: 12.0% (n = 35/291) (p < 0.001) • Adalimumab + MTX: 6.5% (n = 13/199) (p-value NR)
	Month 3	Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall • 200 mg FIL ± csDMARDs: 18.8% (p < 0.001 from baseline)
		Schultz et al.; German multicenter retrospective observational study: • Moderate-to-severe RA	Overall: • 200 mg FIL ± MTX: n = 39/217
	Week 24	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	eRRP vs non-eRRP: • MTX monotherapy: 10.8% vs 10.7% (p = 1.00) • 200 mg FIL + MTX: 19.3% vs 27.1% (p = 0.09) • 100 mg FIL + MTX: 18.2% vs 22.7% (p = 0.31) • Adalimumab + MTX: 8.6% vs 24.7% (p = 0.001)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	eRRP vs non-eRRP: • MTX monotherapy: 16.0% vs 13.1% (p = 0.55) • 200 mg FIL + MTX: 26.8% vs 33.3% (p = 0.20) • 100 mg FIL + MTX: 31.4% vs 27.3% (p = 0.62) • Adalimumab + MTX: 24.2% vs 25.0% (p = 1.00)
	Month 6	Schultz et al.; German multicenter retrospective observational study: • Moderate-to-severe RA	Overall: • 200 mg FIL ± MTX: n = 63/171
SDAI reduction (mean ± SD score)	Week 12	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 10 ± 6 (p < 0.05 from baseline) • 200 mg FIL: 11.2 ± 7.3 (p-value NR) • 200 mg FIL + MTX: 9.2 ± 4.9 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 13.8 ± 8.6 (p-value NR) • 200 mg FIL: 15.8 ± 9.7 (p < 0.05 from baseline) • 200 mg FIL + MTX: 12.3 ± 7.3 (p-value NR)
	Week 24	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 7.5 ± 8 (p < 0.05 from baseline) • 200 mg FIL: 10.5 ± 14.8 (p-value NR) • 200 mg FIL + MTX: 5.5 ± 2.4 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 11.4 ± 7.9 (p-value NR) • 200 mg FIL: 10.8 ± 7.7 (p-value NR) • 200 mg FIL + MTX: 11.8 ± 8.2 (p-value NR)
SDAI remission (proportion of patients)	Week 12	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	4 PPFs: • MTX monotherapy: 2.2% (n = 4/181) (reference group) • 200 mg FIL + MTX: 9.9% (n = 19/191) (p = 0.002) • 100 mg FIL + MTX: 7.4% (n = 14/189) (p = 0.028) • Adalimumab + MTX: 4.8% (n = 6/126) (p-value NR)	< 4 PPFs: • MTX monotherapy: 3.4% (n = 10/294) (reference group) • 200 mg FIL + MTX: 14.8% (n = 42/284) (p < 0.001) • 100 mg FIL + MTX: 10.7% (n = 31/291) (p < 0.001) • Adalimumab + MTX: 8.0% (n = 16/199) (p-value NR)
	Week 24	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	eRRP vs non-eRRP: • MTX monotherapy: 10.0% vs 9.0% (p = 0.840) • 200 mg FIL + MTX: 23.0% vs 28.8% (p = 0.24) • 100 mg FIL + MTX: 17.8% vs 21.4% (p = 0.44) • Adalimumab + MTX: 9.8% vs 24.3% (p = 0.004)
		FINCH 3; Phase 3 RCTs: • Moderate-to-severe MTX-naïve RA	eRRP vs non-eRRP: • MTX monotherapy: 14.2% vs 13.2% (p = 0.88) • 200 mg FIL + MTX: 30.3% vs 34.8% (p = 0.42) • 100 mg FIL + MTX: 33.1% vs 27.3% (p = 0.51) • Adalimumab + MTX: 26.9% vs 25.0% (p = 0.86)
Steroid reduction (mean ± SD steroid dose or median [IQR])	Week 12	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 1.3 ± 2.1 (p < 0.05 from baseline) • 200 mg FIL: 2.6 ± 3.5 (p-value NR) • 200 mg FIL + MTX: 1.6 ± 2.3 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 2.2 ± 3 (p < 0.05 from baseline) • 200 mg FIL: 2.6 ± 3.5 (p-value NR) • 200 mg FIL + MTX: 1.8 ± 2.4 (p-value NR)
	Week 24	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 0.5 ± 1.6 • 200 mg FIL: 0 ± 0 • 200 mg FIL + MTX: 0.7 ± 1.9	bDMARD-IR: • 200 mg FIL ± MTX: 2.4 ± 4.3 • 200 mg FIL: 2.6 ± 5 • 200 mg FIL + MTX: 1.9 ± 3.3
	Month 3	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL + MTX: 0 (0–4) mg (p = 0.0001 vs baseline)
	Month 6	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL + MTX: 0 (0–4) mg (p = 0.0001 vs baseline, p = 0.048 vs Month 3)

Italicized rows indicate real-world studies. DAS28CRP remission defined as < 2.6. DAS28ESR remission defined as < 2.6. CDAI remission defined as ≤ 2.8. SDAI remission defined as ≤ 3.3

Abbreviations: bDMARD biologic disease-modifying anti-rheumatic drug, CDAI Clinical Disease Activity Index, csDMARD conventional synthetic disease-modifying anti-rheumatic drug, DAS28CRP Disease Activity score including 28 joints using C-reactive protein, DAS28ESR Disease Activity score including 28 joints using erythrocyte sedimentation rate, eRRP estimated rapid radiographic progression, FIL filgotinib, GISEA Group for the Study of Early Arthritis, IQR interquartile range, IR inadequate responder, MTX methotrexate, NR not reported, PPF poor prognostic factor, RA rheumatoid arthritis, RCT randomized controlled trial, ReLiFiRa Real-Life Filgotinib in Rheumatoid Arthritis, SD standard deviation, SDAI Simplified Disease Activity Index, tsDMARD targeted synthetic disease-modifying anti-rheumatic drug

*The publication did not specify the type of summary statistic reported; mean ± SD is assumed

In addition to trials directly comparing FIL with adalimumab + MTX and MTX monotherapy, a network meta-analysis (NMA) of RCTs has evaluated the relative effectiveness at Week 12 of 100 mg and 200 mg FIL compared with other JAKis (tofacitinib, baricitinib, upadacitinib), all given in combination with MTX, anchored by placebo + MTX or adalimumab + MTX [19]. Across the network, 200 mg FIL + MTX had the highest probability of achieving CDAI and SDAI remission at Week 12, followed by upadacitinib 15 mg + MTX [19]. Both doses of FIL (100 mg and 200 mg) + MTX achieved significantly better odds ratios for CDAI remission than adalimumab 40 mg + MTX at Week 12, but only 200 mg FIL + MTX achieved significantly better odds ratios for SDAI remission [19]. Regarding ACR70, 200 mg FIL + MTX was ranked as the second-best treatment, after upadacitinib 15 mg + MTX; 200 mg FIL + MTX achieved a significantly better odds ratio than adalimumab 40 mg + MTX [19]. FIL 200 mg + MTX was also among the better performing treatments at Week 24, ranking as the second-best treatment to achieve SDAI remission, following baricitinib 4 mg + MTX; all JAKis included in the network and adalimumab had significantly better odds ratios than placebo + MTX [19]. This aligns with findings from FINCH 1, where a greater proportion of patients receiving FIL 200 mg + MTX (34.1%) compared with adalimumab + MTX (23.7%) achieved DAS28CRP-defined remission at Week 12 (Table 1) [21].

To complement the findings from RCTs, real-world studies have also investigated early outcomes of FIL treatment [22]. An Italian retrospective observational study from 10 rheumatology centers followed 126 patients with RA starting 200 mg FIL ± MTX [23]. Overall remission rates were similar to RCTs; patients who were bDMARD-naïve treated with 200 mg FIL ± MTX had significantly greater reduction in SDAI, CDAI and DAS28CRP scores than IRs to bDMARDs at Week 12 (Table 1) [23]. Reductions from baseline in disease activity were also observed at Week 24 among all patients (Table 1) [23]. In another retrospective multicenter study in Italy that included 82 patients with RA (where approximately 13.4% were classified as difficult to treat due to prior bDMARD failure), almost half of patients achieved DAS28CRP remission at Month 3, which increased to two-thirds by Month 6 (Table 1) [24]. Smaller proportions of patients previously treated with another tsDMARD achieved DAS28CRP remission than those exposed only to bDMARDs (Table 1) [24]. Separately, a German multicenter retrospective chart review of 301 patients with moderate-to-severe RA who initiated FIL + glucocorticoids and/or MTX showed that at Month 3, the proportion of patients achieving CDAI and DAS28CRP remission was similar to that of RCTs; this further increased at Month 6, but no significant differences were observed between FIL doses (100 mg vs 200 mg; Table 1) [25]. Furthermore, interim effectiveness data from a prospective observational study in Europe reported a decrease in DAS28CRP score from baseline to Month 1 (Table 1), as well as reductions in CDAI score (data not shown) [26].

A multicenter retrospective study from Italy following a cohort of patients who had failed ≥ 1 bDMARD observed that among these patients with difficult-to-treat RA who received 200 mg FIL, significant reductions in steroid dosage were observed at Month 3 and Month 6 (Table 1) [22]. Another multicenter Italian study which stratified patients by prior bDMARD exposure showed that FIL allowed difficult-to-treat patients to discontinue or reduce the use of glucocorticoids within 12 weeks (i.e., 3 months), regardless of prior bDMARD exposure (Table 1) [23].

Sustained Effectiveness

FIL has also shown sustained effectiveness beyond Week 24, demonstrating its long-term benefits [20, 27]. In FINCH 1, greater proportions of IRs to MTX with moderate-to-severe RA treated with 200 mg FIL + MTX achieved DAS28CRP remission at Week 52 compared with adalimumab + MTX (Table 2) [21]. However, in patients with moderate RA and inadequate response to MTX, the DAS28CRP remission rates were comparable between either dose of FIL + MTX and adalimumab + MTX at this timepoint (Table 2) [20]. In FINCH 4, which was a long-term extension (LTE) study enrolling patients from the FINCH 1, 2 and 3 parent trials, the proportion of patients achieving CDAI or SDAI remission was sustained for both doses of FIL ± MTX (100 mg and 200 mg) from FINCH 4 baseline to Week 156 across patients of different treatment backgrounds (patients who were MTX-naïve, IRs to MTX and IRs to bDMARDs) (Table 2) [27]. Comparing against other JAKis (tofacitinib, baricitinib, upadacitinib) and adalimumab, an NMA using data from three RCTs demonstrated that 200 mg FIL + MTX had the second-highest probability of achieving CDAI remission at Week 48/52, after upadacitinib 15 mg + MTX [19]. Both FIL and upadacitinib achieved significantly better odds ratios of remission compared with adalimumab 40 mg + MTX [19].

Table 2.

Summary of sustained efficacy and effectiveness findings for FIL from RCTs and real-world studies

Effectiveness measure	Timepoint	Population	Result
DAS28CRP remission (proportion of patients)	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Moderate-to-severe RA: • Adalimumab + MTX: 46% (reference group) • 200 mg FIL + MTX: 54% (p < 0.05) • 100 mg FIL + MTX: 43% (p-value NR)
			Moderate RA: • Adalimumab + MTX: 63.9% • 200 mg FIL + MTX: 62.5% • 100 mg FIL + MTX: 57.0%
	Month 24	FILOSOPHY and PARROTFISH; European prospective observational studies: • Moderate-to-severe RA	b/tsDMARD-exposed: • FIL (100 mg or 200 mg) ± csDMARDs: 62.0% (n = 31/50)	b/tsDMARD-naïve: • FIL (100 mg or 200 mg) ± csDMARDs: 67.6% (n = 23/34)
DAS28ESR remission (proportion of patients)	Month 12	Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall • 200 mg FIL ± csDMARDs: 35.5% (p = 0.002 from baseline)
CDAI remission (proportion of patients)	Month 12	Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall • 200 mg FIL ± csDMARDs: 27.5% (p < 0.001 from baseline)
	Month 24	FILOSOPHY and PARROTFISH; European prospective observational studies: • Moderate-to-severe RA	b/tsDMARD-exposed: • FIL (100 mg or 200 mg) ± csDMARDs: 16.4% (n = 9/55)	b/tsDMARD-naïve: • FIL (100 mg or 200 mg) ± csDMARDs: 35.3% (n = 12/34)
	Week 156	FINCH 4; Phase 3 RCT LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Moderate-to-severe MTX-IR RA: • 200 mg FIL ± MTX (continued FIL exposure): 37.2% • 200 mg FIL ± MTX (de novo FIL exposure): 33.7% • 100 mg FIL ± MTX (continued FIL exposure): 31.0% • 100 mg FIL ± MTX (de novo FIL exposure): 29.3%
			Moderate-to-severe MTX-naïve RA: • 200 mg FIL ± MTX (continued FIL exposure): 43.7% • 200 mg FIL ± MTX (de novo FIL exposure): 41.8% • 100 mg FIL ± MTX (continued FIL exposure): 38.0% • 100 mg FIL ± MTX (de novo FIL exposure): 27.3%
			Moderate-to-severe bDMARD-IR RA: • 200 mg FIL ± csDMARDs (continued FIL exposure): 41.0% • 200 mg FIL ± csDMARDs (de novo FIL exposure): 24.3% • 100 mg FIL ± csDMARDs (continued FIL exposure): 23.9% • 100 mg FIL ± csDMARDs (de novo FIL exposure): 18.9%
SDAI remission (proportion of patients)	Week 156	FINCH 4; Phase 3 RCT LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Moderate-to-severe MTX-IR RA: • 200 mg FIL ± MTX (continued FIL exposure): 26.4% • 200 mg FIL ± MTX (de novo FIL exposure): 21.9% • 100 mg FIL ± MTX (continued FIL exposure): 21.2% • 100 mg FIL ± MTX (de novo FIL exposure): 19.2%
			Moderate-to-severe MTX-naïve RA: • 200 mg FIL ± MTX (continued FIL exposure): 26.8% • 200 mg FIL ± MTX (de novo FIL exposure): 27.0% • 100 mg FIL ± MTX (continued FIL exposure): 24.3% • 100 mg FIL ± MTX (de novo FIL exposure): 20.5%
			Moderate-to-severe bDMARD-IR RA: • 200 mg FIL ± csDMARDs (continued FIL exposure): 22.0% • 200 mg FIL ± csDMARDs (de novo FIL exposure): 13.6% • 100 mg FIL ± csDMARDs (continued FIL exposure): 11.3% • 100 mg FIL ± csDMARDs (de novo FIL exposure): 14.5%

Italic rows indicate real-world studies

Abbreviations: bDMARD biologic disease-modifying anti-rheumatic drug, CDAI Clinical Disease Activity Index, csDMARD conventional synthetic disease-modifying anti-rheumatic drug, DAS28CRP Disease Activity score including 28 joints using C-reactive protein, DAS28ESR Disease Activity score including 28 joints using erythrocyte sedimentation rate, FIL filgotinib, GISEA Group for the Study of Early Arthritis, IR inadequate responder, LTE long-term extension, MTX methotrexate, NR not reported, RA rheumatoid arthritis, RCT randomized controlled trial, SDAI Simplified Disease Activity Index, tsDMARD targeted synthetic disease-modifying anti-rheumatic drug

While FINCH 4 assessed outcomes up to Week 156 as part of an interim analysis (approximately 36 months or 3 years), there are limited published real-world studies reporting outcomes beyond 2 years. An Italian registry study (Group for the Study of Early Arthritis [GISEA]) followed 246 patients, 39% of whom had difficult-to-treat RA, treated with 200 mg FIL (± concomitant glucocorticoids and csDMARDs), and showed significant decreases in disease activity (CDAI and DAS28 using erythrocyte sedimentation rate [DAS28ESR]) between baseline and Month 12 (Table 2) [28]. However, there were no significant differences in disease activity between Month 3 and Month 12 (Table 2) [28].

A German multicenter retrospective study showed that at Month 6, there were no significant differences in CDAI remission rates between FIL doses (100 mg or 200 mg), whether patients received FIL monotherapy or combination therapy, or when stratified by b/tsDMARD history (data not shown) [25]. In comparison, Week 48 (Month 12) data from FINCH 4 among IRs to MTX showed numerically higher proportions of patients achieving DAS28CRP and CDAI remission on 200 mg FIL + MTX than on 100 mg FIL + MTX [13]. Response rates among IRs to bDMARDs were numerically lower than those among IRs to MTX or MTX-naïve patients (data not shown) [13].

However, there remains a lack of longer-term data on the effectiveness of FIL in the real world. Two observational studies are currently ongoing: FILOSOPHY and PARROTFISH are European studies in patients with moderate-to-severe RA receiving FIL as monotherapy or combination therapy for 2 years [29]. The patients enrolled in these two studies had lower baseline disease activity (median CDAI 24.0 and median DAS28CRP 4.5) than those enrolled in the FINCH trials (e.g., FINCH 1: mean CDAI 39.2; median DAS28CRP 5.7) [21, 30]. Interim results from these large observational studies have demonstrated that the proportion of patients achieving CDAI and DAS28CRP remission increased from baseline to Month 1 and 12 and further improved at Month 24 (Table 2) [26, 29]. At Month 24, the proportion of patients achieving CDAI remission was more than double in patients who were naïve to b/tsDMARDs (35.3%) compared with those previously treated with b/tsDMARDs (16.4%; Table 2) [29]. However, proportions of patients achieving DAS28CRP remission at Month 24 were similar between patients who were naïve to b/tsDMARDs and those with prior exposure (62.0–67.6%; Table 2) [29]. Further long-term results are still pending from these ongoing trials.

Symptom Relief

In addition to the effectiveness observed in response rates and decreased disease activity, FIL has also shown effectiveness in providing symptom relief. In post hoc analyses of FINCH 1–3, significant improvements in the visual assessment scale (VAS) pain score were observed with FIL ± csDMARDs as early as Week 2 among IRs to MTX and bDMARDs, which was sustained until study end (Week 24 for FINCH 2 and Week 52 for FINCH 1 and 3) (data not shown) [20, 31]. Likewise, both doses of FIL ± csDMARDs showed significantly greater improvement from baseline in Functional Assessment of Chronic Illness Therapy (FACIT)-fatigue score compared with MTX monotherapy among IRs to MTX and bDMARDs from Week 4, which were sustained until study end (Table 3) [20, 32]. Among patients who were MTX-naïve, both doses of FIL + MTX and 200 mg FIL monotherapy demonstrated significantly greater improvements in fatigue than MTX alone, which was maintained through to Week 52 (Table 3) [32]. Interim results from the FINCH 4 LTE demonstrated reductions in patient assessment of pain with FIL use [27]. Change in patient’s assessment of pain was sustained from baseline (i.e., end of the FINCH parent trial) to Week 156 of FINCH 4 and was numerically similar between both FIL doses (100 mg and 200 mg) ± csDMARDs, as well as between those continuing FIL from the parent trial and those re-randomized to FIL in FINCH 4 (Table 3) [27].

Table 3.

Summary of symptom relief efficacy and effectiveness findings for FIL from RCTs and real-world studies

Effectiveness measure	Timepoint	Population	Result
FACIT-Fatigue score change from baseline (mean ± SE or median [IQR])	Week 4	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • MTX monotherapy: 3.7 ± 0.7 (reference group) • 200 mg FIL + MTX: 6.5 ± 0.7 (p < 0.001) • 100 mg FIL + MTX: 5.9 ± 0.7 (p < 0.001) • Adalimumab + MTX: 5.7 ± 0.7 (p-value NR)
		FINCH 2; Phase 3 RCT: • Moderate-to-severe bDMARD-IR RA	Overall: • csDMARDs: 3.3 ± 1.0 (reference group) • 200 mg FIL + csDMARDs: 6.9 ± 1.0 (p < 0.001) • 100 mg FIL + csDMARDs: 6.6 ± 1.0 (p < 0.001)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall: • MTX monotherapy: 4.4 ± 0.5 (reference group) • 200 mg FIL + MTX: 7.6 ± 0.5 (p < 0.001) • 100 mg FIL + MTX: 6.9 ± 0.6 (p < 0.001) • 200 mg FIL: 7.1 ± 0.6 (p < 0.001)
	Week 24	FINCH 2; Phase 3 RCT: • Moderate-to-severe bDMARD-IR RA	Overall: • csDMARDs: 6.9 ± 1.2 (reference group) • 200 mg FIL + csDMARDs: 11.5 ± 1.1 (p < 0.001) • 100 mg FIL + csDMARDs: 9.0 ± 1.1 (p-value NR)
	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • Adalimumab + MTX: 10.9 ± 0.8 • 200 mg FIL + MTX: 11.7 ± 0.7 • 100 mg FIL + MTX: 11.8 ± 0.7
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall: • MTX monotherapy: 10.4 ± 0.6 (reference group) • 200 mg FIL + MTX: 12.1 ± 0.6 (p < 0.05) • 100 mg FIL + MTX: 11.4 ± 0.8 (p-value NR) • 200 mg FIL: 11.6 ± 0.8 (p-value NR)
	Month 12	FILOSOPHY; European prospective observational study: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 32.0 (21.0–40.0)
VAS patient-reported pain (mean ± SD score or median [IQR])	Week 12	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 4.3 ± 1.4 (p-value NR) • 200 mg FIL: 4.4 ± 1.3 (p < 0.05 from baseline) • 200 mg FIL + MTX: 4.2 ± 1.6 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 4.9 ± 2.2 (p-value NR) • 200 mg FIL: 5.6 ± 2.1 (p-value NR) • 200 mg FIL + MTX: 4.4 ± 2.1 (p-value NR)
	Week 24	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 3.8 ± 2.2 • 200 mg FIL: 4.2 ± 2.7 • 200 mg FIL + MTX: 3.7 ± 2.2	bDMARD-IR: • 200 mg FIL ± MTX: 4.4 ± 2.3 • 200 mg FIL: 4.3 ± 2.2 • 200 mg FIL + MTX: 4.5 ± 2.5
	Month 3	Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall • 200 mg FIL ± csDMARDs: 39.6 ± 29.4 (p < 0.001 from baseline)
		Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 3 ± 2 (p < 0.0001 from baseline)
	Month 6	Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 2 ± 1 (p < 0.0001 from baseline)
	Month 12	Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall • 200 mg FIL ± csDMARDs: 29.5 ± 29.1 (p < 0.001 from baseline)
		FILOSOPHY; European prospective observational study: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 43.5 (18.0–62.0)
VAS patient-reported pain change from baseline (mean ± SD or median [IQR])	Month 3	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL + MTX: 2.5 (1–5) (p = 0.0001 vs baseline)
	Month 6	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL + MTX: 2 (1–3) (p = 0.0001 vs baseline)
	Week 156	FINCH 4; phase 3 RCT LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Moderate-to-severe MTX-IR RA: • 200 mg FIL ± MTX (continued FIL exposure): – 42 • 200 mg FIL ± MTX (de novo FIL exposure): – 40 • 100 mg FIL ± MTX (continued FIL exposure): – 38 • 100 mg FIL ± MTX (de novo FIL exposure): – 33
			Moderate-to-severe MTX-naïve RA: • 200 mg FIL ± MTX (continued FIL exposure): – 43 • 200 mg FIL ± MTX (de novo FIL exposure): – 44 • 100 mg FIL ± MTX (continued FIL exposure): – 41 • 100 mg FIL ± MTX (de novo FIL exposure): – 41
			Moderate-to-severe bDMARD-IR RA: • 200 mg FIL ± csDMARDs (continued FIL exposure): – 39 • 200 mg FIL ± csDMARDs (de novo FIL exposure): – 27 • 100 mg FIL ± csDMARDs (continued FIL exposure): – 35 • 100 mg FIL ± csDMARDs (de novo FIL exposure): – 27
VAS patient-reported fatigue (mean ± SD score)	Week 12	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 4.2 ± 2.1 (p-value NR) • 200 mg FIL: 3.8 ± 1.9 (p < 0.05 from baseline) • 200 mg FIL + MTX: 4.6 ± 2.3 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 4.7 ± 2.3 (p-value NR) • 200 mg FIL: 5.3 ± 2.1 (p-value NR) • 200 mg FIL + MTX: 4.1 ± 2.3 (p-value NR)
	Week 24	La Barbera et al.;* Italian multicenter retrospective observational study: • RA	bDMARD-naïve: • 200 mg FIL ± MTX: 3.4 ± 2.1 (p-value NR) • 200 mg FIL: 3.7 ± 1.5 (p-value NR) • 200 mg FIL + MTX: 3.4 ± 2.4 (p-value NR)	bDMARD-IR: • 200 mg FIL ± MTX: 4.1 ± 2.5 (p-value NR) • 200 mg FIL: 4.3 ± 2.4 (p < 0.05 from baseline) • 200 mg FIL + MTX: 4.3 ± 2.7 (p-value NR)

Italic rows indicate real-world studies

Abbreviations: bDMARD biologic disease-modifying anti-rheumatic drug, csDMARD conventional synthetic disease-modifying anti-rheumatic drug, FACIT Functional Assessment of Chronic Illness Therapy, FIL filgotinib, GISEA Group for the Study of Early Arthritis, IQR interquartile range, IR inadequate responder, LTE long-term extension, MTX methotrexate, NR not reported, RA rheumatoid arthritis, RCT randomized controlled trial, ReLiFiRa Real-Life Filgotinib in Rheumatoid Arthritis, SD standard deviation, SE standard error, VAS visual analog scale

*The publication did not specify the type of summary statistic reported; mean ± SD is assumed

Similar findings for reductions in patient-reported pain and fatigue have also been reported in real-world settings [22–24, 26, 28]. In an Italian registry study, following treatment with 200 mg FIL, the proportion of patients reporting pain of ≤ 20 on the VAS scale (0–100) was significantly larger at Month 3 and Month 12 than at baseline (Table 3) [28]. In another Italian real-world study, FIL monotherapy resulted in significant reductions in VAS pain and fatigue among bDMARD-naïve patients with RA between baseline and Week 12 (Table 3) [23]. When FIL was used in combination with MTX, numerical reductions in VAS pain and fatigue were observed among patients who were bDMARD-naïve and IRs to bDMARDs at Week 12 from baseline, although these were not statistically significant (Table 3) [23]. Extended follow-up at Week 24 showed that in IRs to bDMARDs, FIL monotherapy led to significant reductions in VAS fatigue from baseline (Table 3) [23]. Other retrospective multicenter studies from Italy have also reported significant improvements in VAS pain at Month 3 and Month 6 after treatment with 200 mg FIL (Table 3) [22, 24]. Longer-term findings from FILOSOPHY have shown that improvements in VAS pain and FACIT-fatigue score were observed from Week 1, which were maintained through to Month 12 (Table 3) [26].

Radiographic Outcomes

In addition to clinical outcomes, joint destruction is a key consideration in RA management as structural damage may lead to loss of mobility [33]. Joint-related outcomes include mTSS, which measures radiographic progression based on X-ray imaging and quantifies the bone erosion and joint space narrowing of the hands and feet [34]. As joint space narrowing and bone erosions are generally considered irreversible [35], minimizing radiographic destruction is a key aspect of RA management.

Based on the previously mentioned NMA of RCTs comparing filgotinib to other JAKis (and adalimumab), 200 mg FIL had the highest probability of being the treatment with the greatest improvement in mTSS at Week 48/52, followed by 100 mg FIL [19]. A post hoc analysis of IRs to MTX who had moderate RA from FINCH 1 demonstrated numerically similar radiographic progression with either dose of FIL + MTX compared to adalimumab + MTX at Week 12 and 24; by Week 52, patients on 200 mg FIL + MTX and adalimumab + MTX had numerically similar radiographic progression (Table 4) [20]. Additionally, based on post hoc analyses from FINCH 1 and 3, both doses of FIL + MTX significantly reduced bone erosion at Week 24 from baseline compared with MTX monotherapy, while impact on joint space narrowing was similar between FIL + MTX, MTX monotherapy and adalimumab + MTX (Table 4) [17]. By Week 52, 200 mg FIL + MTX significantly reduced erosion and joint space narrowing from baseline compared with adalimumab + MTX in FINCH 1, while 100 mg FIL + MTX had comparable results to adalimumab + MTX (Table 4) [17]. Further subgroup analyses of FINCH 1 and 3 showed that patients with eRRP had a higher risk of structural progression in subsequent treatment phases versus patients without estimated rapid-radiographic progression (eRRP; data not shown) [17]. Interestingly, only 200 mg FIL + MTX showed consistent suppression of mTSS progression from baseline over 52 weeks, regardless of patients’ eRRP status (Table 4) [17].

Table 4.

Summary of radiographic efficacy and effectiveness findings for FIL from RCTs and real-world studies

Effectiveness measure	Timepoint	Population	Result
mTSS change from baseline (mean or mean ± SD)	Week 12	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • MTX monotherapy: 0.32 (reference group) • 200 mg FIL + MTX: 0.20 (p = 0.28) • 100 mg FIL + MTX: 0.24 (p = 0.46) • Adalimumab + MTX: 0.21 (p-value NR)
	Week 24	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • MTX monotherapy: 0.42 (reference group) • 200 mg FIL + MTX: 0.27 (p = 0.38) • 100 mg FIL + MTX: 0.28 (p = 0.38) • Adalimumab + MTX: 0.19 (p-value NR)
			4 PPFs: • MTX monotherapy: 0.66 (reference group) • 200 mg FIL + MTX: 0.21 (p = 0.002) • 100 mg FIL + MTX: 0.23 (p = 0.004) • Adalimumab + MTX: 0.30 (p-value NR)	< 4 PPFs: • MTX monotherapy: 0.24 (reference group) • 200 mg FIL + MTX: 0.08 (p = 0.059) • 100 mg FIL + MTX: 0.10 (p = 0.100) • Adalimumab + MTX: 0.11 (p-value NR)
			Medium disease activity: • MTX monotherapy: 0.56 (reference group) • 200 mg FIL + MTX: 0.05 (p < 0.05) • 100 mg FIL + MTX: 0.41 (p-value NR) • Adalimumab + MTX: 0.39 (p-value NR)	High disease activity: • MTX monotherapy: 1.12 (reference group) • 200 mg FIL + MTX: -0.26 (p < 0.05) • 100 mg FIL + MTX: 0.72 (p-value NR) • Adalimumab + MTX: 0.39 (p-value NR)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	4 PPFs: • MTX monotherapy: 0.95 (reference group) • 200 mg FIL + MTX: 0.26 (p < 0.05) • 100 mg FIL + MTX: 0.43 (p-value NR) • 200 mg FIL: – 0.18 (p < 0.01)	Overall population: • MTX monotherapy: 0.42 (reference group) • 200 mg FIL + MTX: 0.13 (p-value NR) • 100 mg FIL + MTX: 0.13 (p-value NR) • 200 mg FIL: – 0.13 (p < 0.01)
	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • Adalimumab + MTX: 0.25 • 200 mg FIL + MTX: 0.31 • 100 mg FIL + MTX: 0.52
			eRRP vs non-eRRP: • Adalimumab + MTX: 1.27 ± 5.968 vs 0.23 ± 1.278 (p = 0.896) • 200 mg FIL + MTX: 0.37 ± 2.016 vs 0.14 ± 1.038 (p = 0.137) • 100 mg FIL + MTX: 0.82 ± 2.419 vs 0.34 ± 1.900 (p = 0.024)
			4 PPFs: • Adalimumab + MTX: 0.80 (reference group) • 200 mg FIL + MTX: 0.29 (p = 0.026) • 100 mg FIL + MTX: 0.84 (p = 0.87)	< 4 PPFs: • Adalimumab + MTX: 0.53 (reference group) • 200 mg FIL + MTX: 0.14 (p = 0.002) • 100 mg FIL + MTX: 0.25 (p = 0.022)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	eRRP vs non-eRRP: • MTX monotherapy: 1.01 ± 3.626 vs 0.47 ± 1.807 (p = 0.30) • 200 mg FIL + MTX: 0.34 ± 2.113 vs 0.25 ± 1.070 (p = 0.94) • 100 mg FIL + MTX: 0.31 ± 1.266 vs 0.07 ± 0.716 (p = 0.31) • 200 mg FIL: 0.38 ± 2.207 vs 0.21 ± 1.017 (p = 0.42)
			4 PPFs: • MTX monotherapy: 1.45 (reference group) • 200 mg FIL + MTX: 0.41 (p < 0.01) • 100 mg FIL + MTX: 0.29 (p < 0.01) • 200 mg FIL: 0.40 (p < 0.05)	Overall population: • MTX monotherapy: 0.74 (reference group) • 200 mg FIL + MTX: 0.21 (p < 0.001) • 100 mg FIL + MTX: 0.27 (p < 0.05) • 200 mg FIL: 0.23 (p < 0.01)
Erosion score, change from baseline (mean)	Week 24	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • MTX monotherapy: 0.22 (reference group) • 200 mg FIL + MTX: 0.03 (p < 0.001) • 100 mg FIL + MTX: 0.06 (p < 0.001) • Adalimumab + MTX: 0.06 (p-value NR)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall: • MTX monotherapy: 0.25 (reference group) • 200 mg FIL + MTX: 0.07 (p < 0.05) • 100 mg FIL + MTX: 0.05 (p-value NR) • 200 mg FIL: – 0.01 (p < 0.05)
	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • Adalimumab + MTX: 0.28 (reference group) • 200 mg FIL + MTX: 0.06 (p = 0.001) • 100 mg FIL + MTX: 0.19 (p = 0.18)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall: • MTX monotherapy: 0.33 (reference group) • 200 mg FIL + MTX: 0.07 (p < 0.01) • 100 mg FIL + MTX: 0.10 (p < 0.05) • 200 mg FIL: 0.09 (p < 0.05)
Joint space narrowing score, change from baseline (mean)	Week 24	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • MTX monotherapy: 0.19 (reference group) • 200 mg FIL + MTX: 0.10 (p = 0.11) • 100 mg FIL + MTX: 0.09 (p = 0.078) • Adalimumab + MTX: 0.12 (p-value NR)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall: • MTX monotherapy: 0.17 (reference group) • 200 mg FIL + MTX: 0.06 (p-value NR) • 100 mg FIL + MTX: 0.08 (p-value NR) • 200 mg FIL: – 0.12 (p < 0.05)
	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • Adalimumab + MTX: 0.32 (reference group) • 200 mg FIL + MTX: 0.12 (p = 0.002) • 100 mg FIL + MTX: 0.26 (p = 0.36)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall: • MTX monotherapy: 0.42 (reference group) • 200 mg FIL + MTX: 0.15 (p < 0.01) • 100 mg FIL + MTX: 0.17 (p < 0.05) • 200 mg FIL: 0.13 (p < 0.05)
Tender joint count (median [IQR] or mean ± SD)	Month 3	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL ± MTX: 2 (1–3) (p = 0.0001 from baseline)
		Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 2 ± 3 (p < 0.0001 from baseline)
	Month 6	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL ± MTX: 1 (0–2) (p = 0.0001 from baseline)
		Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 1 ± 2 (p < 0.0001 from baseline)
Swollen joint count (median [IQR] or mean ± SD)	Month 3	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL ± MTX: 1 (1–2) (p = 0.0001 from baseline)
		Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 1 ± 1 (p < 0.0001 from baseline)
	Month 6	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL ± MTX: 0 (0–1) (p = 0.0001 from baseline)
		Raimondo et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 0 ± 1 (p < 0.0001 from baseline)

Italic rows indicate real-world studies

Abbreviations: bDMARD biologic disease-modifying anti-rheumatic drug, eRRP estimated rapid radiographic progression, FIL filgotinib, IQR interquartile range, IR inadequate responder, mTSS modified total Sharp score, MTX methotrexate, NR not reported, PPF poor prognostic factor, RA rheumatoid arthritis, RCT randomized controlled trial, ReLiFiRa Real-Life Filgotinib in Rheumatoid Arthritis, SD standard deviation

In addition to eRRP status, clinicians may also consider other existing PPFs, such as baseline bone erosions when prioritizing treatment to manage radiographic progression [18, 36]. According to data from FINCH 1, among IRs to MTX with 4 PPFs, FIL (100 mg and 200 mg) + MTX significantly reduced mTSS compared with MTX monotherapy (Table 4) [18, 36]. However, only 200 mg FIL + MTX, not 100 mg, was associated with consistently higher proportions of patients with no radiographic progression among patients with 4 PPFs (Table 4) [18]. Similar trends were observed in patients who were MTX-naïve from FINCH 3 (Table 4) [36].

While FIL’s effectiveness in reducing disease activity may play a role in reduced joint destruction, a post hoc analysis suggested an uncoupling of disease activity and joint destruction for 200 mg FIL + MTX among IRs to MTX who had moderate-to-severe RA [37]. Among a study investigating 200 mg FIL + MTX, 100 mg FIL + MTX and adalimumab + MTX, 200 mg FIL + MTX was the only treatment that significantly inhibited joint damage compared with MTX monotherapy in patients with medium or high disease activity (Table 4) [37]. Notably, this uncoupling was not observed in patients treated with 100 mg FIL + MTX, further suggesting a potential dose-dependent response between FIL and joint destruction [37].

Published reports of real-world joint outcomes among patients treated with FIL are limited. However, improved radiographic progression and bone erosion have been observed with 200 mg FIL + MTX in a small observational cohort in Japan [38]. This aligns with findings from RCTs showing reduced radiographic progression among patients using FIL. Among the ten patients assessed for bone erosion in their hands and feet, all had extensive bone erosion at baseline that resolved with 200 mg FIL + MTX use in almost all patients by Month 12 (data not shown) [38]. Further Italian retrospective studies also demonstrated that 200 mg FIL significantly reduced tender and swollen joint counts among patients with RA at Month 3 and Month 6 compared with baseline measurements (Table 4) [22, 24]. Given the sustained improvements in joint destruction, the use of FIL ± MTX at 200 mg may be an optimal long-term treatment option for the management of structural changes in patients with RA.

Functional Endpoints

Structural joint damage is closely linked to functional endpoints, particularly in the long term, as joint destruction can lead to loss of physical function and limitations in daily activities [33]. FIL + MTX has shown clinically meaningful improvements in patient functional status up to Week 52 compared with MTX monotherapy or adalimumab + MTX, as observed in the DARWIN and FINCH trials [13].

Post hoc analyses of FINCH 1–3 showed that among patients who were MTX-naïve, or IRs to bDMARDs or csDMARDs enrolled in each of the trials, improvements in Health Assessment Questionnaire Disability Index (HAQ-DI) from baseline were greater for all FIL doses ± MTX compared with MTX alone at Week 4 through to Week 24; by Week 52, only 200 mg FIL as monotherapy or combination therapy had greater improvements than MTX monotherapy among patients who were MTX-naïve (Table 5) [32]. Similarly, among IRs to MTX, improvements in HAQ-DI were significantly greater at Week 4 through to Week 24 with 200 mg or 100 mg FIL + MTX compared with MTX monotherapy (Table 5) [32]. Compared with adalimumab + MTX, 200 mg FIL + MTX resulted in similar improvements to HAQ-DI at Week 4 and 24; however, by Week 52, improvements in HAQ-DI were numerically greater with 200 mg FIL + MTX than adalimumab + MTX (Table 5) [32].

Table 5.

Summary of functional efficacy and effectiveness findings for FIL from RCTs and real-world studies

Effectiveness measure	Timepoint	Population	Result
HAQ-DI change from baseline (mean, mean ± SD or median [IQR])	Week 4	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall:* • MTX monotherapy: – 0.25 ± 0.04 (reference group) • 200 mg FIL + MTX: – 0.44 ± 0.04 (p < 0.001) • 100 mg FIL + MTX: – 0.35 ± 0.04 (p < 0.001) • Adalimumab + MTX: – 0.40 ± 0.04 (p-value NR)
		FINCH 2; Phase 3 RCT: • Moderate-to-severe bDMARD-IR RA	Overall:* • csDMARDs: – 0.13 ± 0.05 (reference group) • 200 mg FIL + csDMARD: – 0.35 ± 0.05 (p < 0.001) • 100 mg FIL + csDMARD: – 0.28 ± 0.05 (p < 0.01)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall:* • MTX monotherapy: – 0.35 ± 0.03 (reference group) • 200 mg FIL + MTX: – 0.64 ± 0.03 (p < 0.001) • 100 mg FIL + MTX: –0.50 ± 0.04 (p < 0.001) • 200 mg FIL: – 0.56 ± 0.04 (p < 0.001)
	Month 3	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL ± MTX: 0.25 (0–0.5) (p < 0.0001 from baseline)
		Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall: • 200 mg FIL ± csDMARDs: 0.9 ± 0.7 (p < 0.05 from baseline)
	Week 24	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall:* • MTX monotherapy: – 0.55 ± 0.04 (reference group) • 200 mg FIL + MTX: – 0.82 ± 0.04 (p < 0.001) • 100 mg FIL + MTX: – 0.74 ± 0.04 (p < 0.001) • Adalimumab + MTX: – 0.76 ± 0.04 (p-value NR)
		FINCH 2; Phase 3 RCT: • Moderate-to-severe bDMARD-IR RA	Overall:* • csDMARDs: – 0.27 ± 0.06 (reference group) • 200 mg FIL + csDMARD: – 0.63 ± 0.06 (p < 0.001) • 100 mg FIL + csDMARD: – 0.49 ± 0.06 (p < 0.01)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall:* • MTX monotherapy: – 0.81 ± 0.03 (reference group) • 200 mg FIL + MTX: – 1.00 ± 0.03 (p < 0.001) • 100 mg FIL + MTX: – 0.94 ± 0.04 (p < 0.01) • 200 mg FIL: – 0.91 ± 0.04 (p < 0.05)
	Month 6	Benucci et al.; ReLiFiRa multicenter retrospective Italian observational study: • RA with prior ≥ 1 bDMARD failure	Overall: • 200 mg FIL ± MTX: 0.5 (0–0.5) (p < 0.0001 from baseline)
	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall:* • Adalimumab + MTX: – 0.82 ± 0.05 (reference group) • 200 mg FIL + MTX: – 0.90 ± 0.04 (p < 0.05) • 100 mg FIL + MTX: – 0.82 ± 0.04 (p-value NR)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall:* • MTX monotherapy: – 0.87 ± 0.04 (reference group) • 200 mg FIL + MTX: – 1.03 ± 0.04 (p < 0.001) • 100 mg FIL + MTX: – 0.96 ± 0.05 (p-value NR) • 200 mg FIL: – 0.98 ± 0.05 (p < 0.05)
	Month 12	Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall: • 200 mg FIL ± csDMARDs: 0.8 ± 0.9 (p < 0.01 from baseline)
	Week 156	FINCH 4; Phase 3 RCT LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Moderate-to-severe MTX-IR RA: • 200 mg FIL ± MTX (continued FIL exposure): – 0.94 • 200 mg FIL ± MTX (de novo FIL exposure): – 0.81 • 100 mg FIL ± MTX (continued FIL exposure): – 0.78 • 100 mg FIL ± MTX (de novo FIL exposure): – 0.77
			Moderate-to-severe MTX-naïve RA: • 200 mg FIL ± MTX (continued FIL exposure): – 1.00 • 200 mg FIL ± MTX (de novo FIL exposure): – 0.95 • 100 mg FIL ± MTX (continued FIL exposure): – 0.95 • 100 mg FIL ± MTX (de novo FIL exposure): – 0.94
			Moderate-to-severe bDMARD-IR RA: • 200 mg FIL ± csDMARDs (continued FIL exposure): – 0.76 • 200 mg FIL ± csDMARDs (de novo FIL exposure): – 0.67 • 100 mg FIL ± csDMARDs (continued FIL exposure): – 0.70 • 100 mg FIL ± csDMARDs (de novo FIL exposure): – 0.58

Italic rows indicate real-world studies

Abbreviations: bDMARD biologic disease-modifying anti-rheumatic drug, csDMARD conventional synthetic disease-modifying anti-rheumatic drug, FIL filgotinib, GISEA Group for the Study of Early Arthritis, HAQ-DI Health Assessment Questionnaire Disability Index, IQR interquartile range, IR inadequate responder, LTE long-term extension, MTX methotrexate, NR not reported, RA rheumatoid arthritis, RCT randomized controlled trial, ReLiFiRa Real-Life Filgotinib in Rheumatoid Arthritis, SD standard deviation, SE standard error

*indicate where publications reported mean ± SE instead of mean ± SD

Significantly greater improvement in physical function by FIL + MTX versus adalimumab + MTX was also observed at Week 52 in IRs to MTX with 4 PPFs [18], although in patients with < 4 PPFs, there was no significant difference between FIL + MTX and adalimumab + MTX in HAQ-DI reduction (data not shown) [18].

Improvements in functional outcomes on FIL can be sustained in the long term, as demonstrated in the interim report of FINCH 4. Change from baseline in HAQ-DI remained constant until Week 156 and was similar between FIL doses (100 mg and 200 mg) ± csDMARDs regardless of prior MTX exposure (Table 5) [27]. Among IRs to bDMARDs, HAQ-DI remained stable until Week 156 among those who continued to receive FIL from the parent trial to FINCH 4; among those who were re-randomized to FIL, HAQ-DI improved from FINCH 4 baseline to Week 156 (Table 5) [27].

Based on evidence from real-world settings and RCTs, FIL has demonstrated effectiveness in improving functional outcomes in patients with RA, regardless of prior treatment history or patient risk factors. Corroborating the positive effects on functional outcomes observed in RCTs, a real-world Italian registry study following patients with RA reported a significant reduction in HAQ-DI with 200 mg FIL ± csDMARDs, which was significant at Month 3 and Month 12 (Week 52; Table 5) [28]. Another retrospective multicenter study from Italy showed improvements in HAQ-DI from baseline to Month 3, which was sustained through to Month 6 with 200 mg FIL treatment (Table 5) [22].

Safety and Tolerability of Filgotinib

In addition to effectiveness, the safety and tolerability of FIL should be carefully assessed. As FIL suppresses the inflammatory pathway, patients may be more vulnerable to specific adverse events (AEs) of interest, such as infections [39]. After a safety study of tofacitinib identified an increased risk of malignancy in patients receiving tofacitinib, this prompted new recommendations, as well as warning and precautionary measures, by regulatory agencies for all JAKis, including FIL [39]. As such, these key AEs are of particular interest for understanding FIL tolerability.

Overall, the safety profile of FIL ± MTX has been well established based on findings from the DARWIN and FINCH trials and their LTEs. Rates of AEs, such as herpes zoster (HZ), serious infections, major adverse cardiovascular events and malignancy, were generally low in RCTs, and were similar to those treated with adalimumab (data not shown) [13]. The FINCH 4 LTE study showed that safety profiles can differ based on patients’ treatment histories; for example, the numbers of serious AEs, such as malignancies and infections, were higher among IRs to bDMARDs than IRs to MTX or patients naïve to MTX (data not shown) [13]. Additionally, there were some differences in the safety profile by duration of FIL treatment. Among patients who were MTX-naïve, rates of AEs of interest were numerically higher in those who were on FIL for a longer duration (i.e., continued FIL from the parent trial through the LTE) than in those who were re-randomized from MTX monotherapy to FIL (data not shown) [13]. However, among IRs to bDMARDs, the rates of AEs of interest were similar between the parent trial and the LTE (data not shown) [13].

While the duration of FIL and treatment history may have affected the safety profile of FIL, the FIL dose prescribed had little effect. The exposure-adjusted incidence rates (EAIRs) of TEAEs were similar between 100 and 200 mg FIL ± csDMARDs in the overall population [27]. Furthermore, an integrated safety analysis of all three DARWIN and all four FINCH trials over a treatment duration of up to 8.3 years showed that the incidence rates of overall grade ≥ 3 TEAEs and serious TEAEs were similar between 100 and 200 mg FIL (Table 6) [39]. EAIRs for malignancies (non-melanoma skin cancer [NMSC] and other malignancies) were similar between 100 mg FIL and 200 mg FIL ± csDMARDs and stable over time (Table 6) [39]. Likewise, EAIR/100 patient-years of exposure (PYE) of HZ were comparable between FIL ± csDMARDs doses (Table 6) [39]. While EAIRs of serious infections were also comparable between FIL doses and remained stable over time (Table 6) [39], 200 mg FIL ± csDMARDs had a numerically lower incidence of overall infections in the long term [13].

Table 6.

Summary of safety findings for FIL from RCTs and real-world studies

Safety measure	Timepoint/duration	Study design/population	Result
TEAEs (proportion of patients or incidence rates)	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 71.6% (n = 111/155) vs 75.3% (n = 241/320) • 100 mg FIL + MTX: 69.0% (n = 107/155) vs 74.8% (n = 243/325)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 77.9% (n = 201/258) vs 74.1% (n = 117/158) • 100 mg FIL + MTX: 84.4% (n = 114/135) vs 69.4% (n = 50/72)
	Month 24	FILOSOPHY and PARROTFISH; European prospective observational studies: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 53.5% (n = 700/1309)
	Week 156	FINCH 4; Phase 3 RCT LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± csDMARDs: 86.47% (n = 1323/1530), 93.7 EAIR/100 PYE (95% CI: 88.7, 98.9) • 100 mg FIL ± csDMARDs: 84.82% (n = 1017/1199), 91.3 EAIR/100 PYE (95% CI: 85.8, 97.1)  < 65 years old vs ≥ 65 years old: • FIL (200 mg vs 100 mg) ± csDMARDs: 97.9 EAIR/100 PYE (95% CI: 94.2, 101.7) vs 120.4 EAIR/100 PYE (95% CI: 111.1, 130.2)
Grade ≥ 3 TEAE (proportion of patients or incidence rates)	8.3 years	DARWIN 1–3, FINCH 1–4; Phase 2 and 3 RCTs and LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± MTX: 22.6% (n = 513/2267), 7.1 EAIR/100 PYE (95% CI: 6.5, 7.8) • 100 mg FIL ± MTX: 20.6% (n = 340/1647), 8.5 EAIR/100 PYE (95% CI: 7.6, 9.4)
Serious TEAE (number or proportion of patients or incidence rates)	Month 12	FILOSOPHY; European prospective observational study: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: n = 20/155
	Month 24	FILOSOPHY and PARROTFISH; European prospective observational studies: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 9.1% (n = 119/1309)
	8.3 years	DARWIN 1–3, FINCH 1–4; Phase 2 and 3 RCTs and LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± MTX: 19.9% (n = 450/2267), 6.1 EAIR/100 PYE (95% CI: 5.5, 6.7) • 100 mg FIL ± MTX: 17.8% (n = 293/1647), 7.1 EAIR/100 PYE (95% CI: 6.3, 8.0)  < 65 years old vs ≥ 65 years old: • FIL (200 mg vs 100 mg) ± csDMARDs: 5.5 EAIR/100 PYE (95% CI: 5.0, 6.0) vs 11.3 EAIR/100 PYE (95% CI: 9.9, 13.0)
NMSC malignancy TEAE (number or proportion of patients or incidence rates)	Month 12	FILOSOPHY; European prospective observational study: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: n = 4/155
	8.3 years	DARWIN 1–3, FINCH 1–4; Phase 2 and 3 RCTs and LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± MTX: 1.2% (n = 27/2267), 0.3 EAIR/100 PYE (95% CI: 0.2, 0.5) • 100 mg FIL ± MTX: 0.5% (n = 9/1647), 0.2 EAIR/100 PYE (95% CI: 0.1, 0.4)  < 65 years old vs ≥ 65 years old: • 200 mg FIL ± csDMARDs: 0.5% (n = 10/1860), 0.2 EAIR/100 PYE (95% CI: 0.1, 0.3) vs 4.2% (n = 17/407), 1.4 EAIR/100 PYE (95% CI: 0.8, 2.2) • 100 mg FIL ± csDMARDs: 0.4% (n = 5/1321), 0.1 EAIR/100 PYE (95% CI 0.0, 0.3) vs 1.2% (n = 4/326), 0.4 EAIR/100 PYE (95% CI: 0.1, 1.1)
Malignancy TEAE (excluding NMSC; number or proportion of patients or incidence rates)	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	4PPFs vs < 4 PPFs: • 200 mg FIL + MTX: 0.0% (n = 0/191) vs 0.7% (n = 2/284) • 100 mg FIL + MTX: 0.0% (n = 0/189) vs 0.7% (n = 2/291)
	Month 12	FILOSOPHY; European prospective observational study: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: n = 4/155
	8.3 years	DARWIN 1–3, FINCH 1–4; Phase 2 and 3 RCTs and LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± MTX: 2.5% (n = 57/2267), 0.7 EAIR/100 PYE (95% CI: 0.5, 0.9) • 100 mg FIL ± MTX: 1.8% (n = 30/1647), 0.7 EAIR/100 PYE (95% CI: 0.4, 0.9)  < 65 years old vs ≥ 65 years old: • 200 mg FIL ± csDMARDs: 1.7% (n = 32/1860), 0.5 EAIR/100 PYE (95% CI: 0.3, 0.7) vs 6.1% (n = 25/407), 2.0 EAIR/100 PYE (95% CI: 1.3, 2.9) • 100 mg FIL ± csDMARDs: 1.6% (n = 21/1321), 0.6 EAIR/100 PYE (95% CI 0.4, 0.9) vs 2.8% (n = 9/326), 1.0 EAIR/100 PYE (95% CI: 0.5, 1.9)
HZ infection TEAE (number or proportion of patients or incidence rates)	Week 24	La Barbera et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL ± MTX: 0% (n = 0/126)
	Week 52	FINCH 1; phase 3 RCT: • Moderate-to-severe MTX-IR RA	4PPFs vs < 4 PPFs: • 200 mg FIL + MTX: 0.5% (n = 1/191) vs 1.8% (n = 5/284) • 100 mg FIL + MTX: 0.5% (n = 1/189) vs 1.0% (n = 3/291)
	Month 12	FILOSOPHY; European prospective observational study: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: n = 3/155
	8.3 years	DARWIN 1–3, FINCH 1–4; Phase 2 and 3 RCTs and LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± MTX: 5.0% (n = 114/2267), 1.5 EAIR/100 PYE (95% CI: 1.2, 1.8) • 100 mg FIL ± MTX: 3.0% (n = 49/1647), 1.1 EAIR/100 PYE (95% CI: 0.8, 1.5)
Infection TEAE (number or proportion of patients)	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 42.6% (n = 66/155) vs 43.8% (n = 140/320) • 100 mg FIL + MTX: 36.8% (n = 57/155) vs 42.2% (n = 137/243)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 41.1% (n = 106/258) vs 26.6% (n = 42/158) • 100 mg FIL + MTX: 40.0% (n = 54/135) vs 30.6% (n = 22/72) • 200 mg FIL: 33.6% (n = 47/140) vs 40.0% (n = 28/70)
Serious infection TEAE (number or proportion of patients or incidence rates)	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	4PPFs vs < 4 PPFs: • 200 mg FIL + MTX: 2.6% (n = 5/191) vs 2.8% (n = 8/284) • 100 mg FIL + MTX: 2.1% (n = 4/189) vs 3.1% (n = 9/291)
	8.3 years	DARWIN 1–3, FINCH 1–4; Phase 2 and 3 RCTs and LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± MTX: 6.6% (n = 149/2267), 1.9 EAIR/100 PYE (95% CI: 1.6, 2.2) • 100 mg FIL ± MTX: 5.9% (n = 97/1647), 2.2 EAIR/100 PYE (95% CI: 1.8, 2.7)  < 65 years old vs ≥ 65 years old: • 200 mg FIL ± csDMARDs: 6.1% (n = 113/1,860), 1.7 EAIR/100 PYE (95% CI: 1.4, 2.0) vs 8.8% (n = 36/407), 2.9 EAIR/100 PYE (95% CI: 2.0, 4.0) • 100 mg FIL ± csDMARDs: 5.4% (n = 71/1321), 2.0 EAIR/100 PYE (95% CI 1.6, 2.5) vs 8.0% (n = 26/326), 2.9 EAIR/100 PYE (95% CI: 1.9, 4.3)
Hematological TEAE (mild worsening of leukopenia, neutropenia, lymphopenia, thrombocytopenia, anemia; number of patients)	Week 12	La Barbera et al.; Italian multicenter retrospective observational study: • RA	Overall: • 200 mg FIL monotherapy: n = 16/62
	Week 24		Overall: • 200 mg FIL monotherapy: n = 4/62 • 200 mg FIL + MTX: n = 5/64

Italic rows indicate real-world studies

Abbreviations: bDMARD biologic disease-modifying anti-rheumatic drug, CI confidence interval, csDMARD conventional synthetic disease-modifying anti-rheumatic drug, EAIR exposure-adjusted incidence rate, eRRP estimated rapid radiographic progression, FIL filgotinib, HZ herpes zoster, IR inadequate responder, LTE long-term extension, MTX methotrexate, NMSC non-melanoma skin cancer, NR not reported, PPF poor prognostic factor, PYE patient-years of exposure, RA rheumatoid arthritis, RCT randomized controlled trial, TEAE treatment-emergent adverse event

Additional subgroup analyses of older patients (≥ 65 years) enrolled in the DARWIN and FINCH trials have been conducted as this group has been shown to be at increased risk of safety events [39]. EAIRs for any TEAEs and serious TEAEs were generally higher in older patients than in those aged < 65 years (Table 6) [39]. Aligning with the overall population, EAIRs of serious infections were similar for 100 mg FIL ± csDMARDs and 200 mg FIL ± csDMARDs in patients aged ≥ 65 years (Table 6) [39]. Additionally, numerically lower EAIRs were observed with 100 mg FIL ± csDMARDs than with 200 mg FIL ± csDMARDs for both NMSC and other malignancies in patients aged ≥ 65 years (Table 6) [39]. This suggests a dose-related trend towards increased incidence specifically among older patients [39]. Such patients are more susceptible to renal impairment [40], and as FIL is metabolized and excreted in the urine via the kidney, the current usage guidelines and the package insert in Japan suggest a starting dose of 100 mg in patients with moderate or severe renal impairment (15 ≤ eGFR < 60) [11, 14]. Therefore, older patients should be screened for renal impairment before initiating FIL and during FIL treatment. Based on European guidelines, patients with inadequate response to MTX may consider JAKis as long as risk factors are considered [6]; a lower starting dose of 100 mg FIL is recommended for patients at risk of venous thromboembolism, major adverse cardiovascular events and malignancy [41]. FIL may be increased to 200 mg in cases of insufficient disease control, but the lowest effective dose is recommended for long-term RA management [41].

Similar dose-related trends in safety events to those observed in older age were also reported in post hoc analyses of FINCH 1 and 3 by eRRP status [17]. While overall TEAEs were generally similar between treatments regardless of eRRP status, 200 mg FIL monotherapy was the only treatment arm that did not show numerically higher rates of infection among patients with eRRP than in those without eRRP (Table 6) [17].

In further analyses of safety among patient subgroups, post hoc analyses showed that IRs to MTX with 4 PPFs did not experience a higher incidence of any particular TEAE with FIL ± MTX than those with < 4 PPFs [18, 36]. Interestingly, incidences of TEAEs, such as serious infections, HZ infection and malignancy, were lower among IRs to MTX with 4 PPFs than in those with < 4 PPFs (Table 6) [36]. Contrastingly, among patients who were MTX-naïve, infection rates were generally higher in patients with 4 PPFs or with eRRP [36].

While real-world studies reporting on long-term safety are limited, FIL has shown a comparable real-world short-term safety profile to that established in RCTs. In a multicenter Italian study of 126 patients with RA, no cases of HZ infection were identified (Table 6) [23]. However, there were some changes in hematological parameters in patients treated with FIL monotherapy that were deemed insignificant at Week 12 and Week 24 (Table 6) [23]. This is broadly aligned with the recently published EULAR consensus, which included a statement on how JAK2-inhibiting JAKis may lead to anemia, while caveating that JAK1-inhibiting FIL demonstrated improvements in anemia in the FINCH and DARWIN trials [42]. Finally, interim findings from European non-interventional studies (FILOSOPHY and PARROTFISH) have also shown that the safety profiles of 200 mg FIL in patients with moderate-to-severe RA were consistent with those of RCTs at Months 12 and Months 24 (Table 6) [26, 43].

Persistence on Filgotinib

Patients enrolled in RCTs typically discontinue study drugs because of inefficacy or safety, such as serious adverse events [44]; comparatively, as real-world studies are conducted in less controlled settings, additional reasons and considerations behind drug discontinuation can be elucidated [25, 45]. Treatment persistence can be used as a surrogate endpoint for patient outcomes and lead to a greater understanding of the potential impacts on real-world treatment effectiveness and patient preferences [45].

Overall, across all of the DARWIN and FINCH trials, the EAIRs of TEAEs leading to premature treatment discontinuation were numerically lower in 100 mg FIL ± csDMARDs than in 200 mg FIL ± csDMARDs (Table 7) [39]. Notably, EAIRs for TEAEs leading to treatment discontinuation were generally higher in patients aged ≥ 65 years than in those < 65 years (Table 7) [39].

Table 7.

Summary of persistence on FIL from RCTs and real-world studies

Persistence measure	Timepoint/follow-up	Population	Result
Completion rate of RCTs (proportion of patients)	Week 24	FINCH 2; Phase 3 RCT: • Moderate-to-severe bDMARD-IR RA	Overall: • FIL (100 mg or 200 mg) + csDMARDs: 85.0%
	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	Overall: • FIL (100 mg or 200 mg) + MTX: 86.4%
	Week 52	FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	Overall: • FIL (100 mg or 200 mg) ± MTX: 82.1%
	Week 156	FINCH 4; Phase 3 RCT LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 63%
	Week 204	DARWIN 3; Phase 2 RCT LTE: • Moderate-to-severe MTX-IR RA	Overall: • FIL (100 mg or 200 mg) ± MTX: 54.3%
TEAEs leading to premature treatment discontinuation (number or proportion of patients or incidence rates)	Week 52	FINCH 1; Phase 3 RCT: • Moderate-to-severe MTX-IR RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 6.5% (n = 10/155) vs 5.0% (n = 16/320) • 100 mg FIL + MTX: 2.6% (n = 4/155) vs 3.4% (n = 11/325)
		FINCH 3; Phase 3 RCT: • Moderate-to-severe MTX-naïve RA	eRRP vs non-eRRP: • 200 mg FIL + MTX: 7.4% (n = 19/258) vs 5.7% (9/158) • 100 mg FIL + MTX: 8.1% (n = 11/135) vs 2.8% (2/72) • 200 mg FIL monotherapy: 2.1% (n = 3/140) vs 2.9% (2/70)
	Month 24	FILOSOPHY and PARROTFISH; European prospective observational studies: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 9.7% (n = 127/1309)
	Week 156	FINCH 4; Phase 3 RCT LTE: • Moderate-to-severe MTX-IR, MTX-naïve and bDMARD-IR RA	Overall: • 200 mg FIL ± csDMARDs: 10.72% (n = 164/1530), 3.6 EAIR/100 PYE (95% CI: 3.1, 4.2) • 100 mg FIL ± csDMARDs: 10.76% (n = 129/1199), 3.6 EAIR/100 PYE (95% CI: 3.0, 4.3)
	8.3 years	DARWIN 1–3, FINCH 1–4; Phase 2 and 3 RCTs and LTE: • Moderate-to-severe MTX-IR, MTX	Overall: • 200 mg FIL ± MTX: 17.6% (n = 398/2267), 5.0 EAIR/100 PYE (95% CI: 4.6, 5.6) • 100 mg FIL ± MTX: 11.1% (n = 182/1647), 4.0 EAIR/100 PYE (95% CI: 3.5, 4.7)
			≥ 65 years vs < 65 years: • FIL (100 mg or 200 mg) + MTX: 7.3 EAIR/100 PYE (95% CI: 6.2, 8.6) vs 4.1 EAIR/100 PYE (95% CI: 3.7, 4.5)
Discontinuation rate (proportion of patients)	Month 6	Schultz et al.; German multicenter retrospective observational study: • Moderate-to-severe RA	Overall: • 200 mg FIL ± MTX: 12.2% (n = 33/301)
	Month 12	Schultz et al.; German multicenter retrospective observational study: • Moderate-to-severe RA	Overall: • 200 mg FIL ± MTX: 18.4% (n = 41/301)
Persistence (proportion of patients [95% CI] or mean drug survival time)	Month 6	FILOSOPHY and PARROTFISH; European prospective observational studies: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 84.9% (82.8–86.8%)
	Month 12	FILOSOPHY and PARROTFISH; European prospective observational studies: • Moderate-to-severe RA	Overall: • FIL (100 mg or 200 mg) ± csDMARDs: 74.6% (72.1–76.9%)
		Fornaro et al.; Retrospective Italian GISEA registry observational study: • RA	Overall: • 200 mg FIL + csDMARDs: 71% persistence, mean drug survival time 17 months (95% CI: 14, 20) • 200 mg FIL: 75% persistence, mean drug survival time 21 months (95% CI: 16, 25)
			Difficult-to-treat RA vs not-difficult-to-treat RA: • 200 mg FIL ± csDMARDs: 65.9% persistence, mean drug survival time 18 months (95% CI: 13, 23) vs 79.5% persistence, mean drug survival time 20 months (95% CI: 17, 22) p = 0.03
	1 Year	Takeuchi et al.; Japanese retrospective claims database study: • RA	csDMARD-IR: • FIL (100 mg or 200 mg): 67%
			bDMARD-IR: • FIL (100 mg or 200 mg): 66%
	2 Year	Takeuchi et al.; Japanese retrospective claims database study: • RA	csDMARD-IR: • FIL (100 mg or 200 mg): 62% 200 mg FIL vs 100 mg FIL in csDMARD-IR: • HR: 0.59 (95% CI: 0.30, 1.19)  ≥ 65 years old vs < 65 years old in csDMARD-IR: • HR: 2.24 (95% CI: 0.90, 5.60)
			bDMARD-IR: • FIL (100 mg or 200 mg): 64% 200 mg FIL vs 100 mg FIL in bDMARD-IR: • HR: 0.72 (95% CI: 0.29, 1.76)  ≥ 65 years old vs < 65 years old in bDMARD-IR: • HR: 1.62 (95% CI: 0.59, 4.45)

Italic rows indicate real-world studies

Abbreviations: bDMARD biologic disease-modifying anti-rheumatic drug, CI confidence interval, csDMARD conventional synthetic disease-modifying anti-rheumatic drug, EAIR exposure-adjusted incidence rate, eRRP estimated rapid-radiographic progression, FIL filgotinib, GISEA Group for the Study of Early Arthritis, HR hazard ratio, IR inadequate responder, LTE long-term extension, MTX methotrexate, NR not reported, PYE patient-years of exposure, RA rheumatoid arthritis, RCT randomized controlled trial, TEAE treatment-emergent adverse event

The completion rates of FIL treatments from the FINCH parent trials were generally high (> 80%; Table 7) [13]. In both IRs to MTX and patients naïve to MTX, TEAEs leading to discontinuation between treatment arms were similar regardless of eRRP status (Table 7) [17]. In FINCH 4, the completion rate was much lower at Week 156 than in the parent trials, where approximately 60% of patients remained on the study drug during the interim analysis (Table 7) [27]. The proportion of patients experiencing TEAEs that led to discontinuation were similar between FIL doses (100 mg and 200 mg) ± csDMARDs (Table 7) [27]. Comparatively, similar proportions of patients completed FIL ± MTX treatment in DARWIN 3, which was also lower than the completion rate of its parent trials DARWIN 1 and 2 (Table 7) [13].

While measures of adherence and persistence are typically limited to discontinuation due to TEAEs in RCTs, data from observational studies can provide additional insights behind FIL discontinuation in the real world. Patients on FIL generally demonstrated high treatment persistence for up to 1 year (75–82%) in several real-world cohorts [25, 28, 46].

An Italian registry study showed that persistence rates differ by severity of RA; patients with difficult-to-treat RA (defined as failure of ≥ 2 b/tsDMARDs, signs suggesting active/progressive disease and patient VAS pain and/or patient/physician global assessment > 20) had significantly lower persistence than other patients with RA (Table 7) [28]. Similar to findings in RCTs, discontinuation rates of FIL ± MTX tended to increase over time; among patients with moderate-to-severe RA, the discontinuation rate of FIL ± MTX numerically increased (~ 4%) from Month 6 to Month 12 (Table 7) [25].

Among real-world studies, the most cited reason for discontinuation was lack of effectiveness [25, 28]. Other reasons for discontinuation included AEs, loss of efficacy and lack of adherence [23, 25, 28]. Among those who had discontinued FIL because of lack of efficacy in a German real-world study, the majority (84.6%) had previously received b/tsDMARDs, and approximately half (57.7%) had received prior JAKis [25].

Notably, no significant differences in FIL discontinuation rates were observed between 100 mg FIL and 200 mg FIL [25, 45] or by FIL regimen (monotherapy vs combination therapy) (Table 7) [25, 28]. Additionally, FIL had comparable persistence to other JAKis and bDMARDs among patients with moderate-to-severe RA who were IRs to csDMARDs or bDMARDs (data not shown) [45]. The 1- and 2-year persistence rates of FIL were similar between IRs to csDMARDs and IRs to bDMARDs and were similar to those reported in clinical trials (Table 7) [45]. Further subgroup analysis showed that patients aged ≥ 65 years were more likely to discontinue FIL than those aged < 65 years, although this difference was not statistically significant (Table 7) [45].

Longer term real-world persistence of FIL among patients with moderate-to-severe RA has been reported in an interim analysis of the FILOSOPHY and PARROTFISH observational studies in Europe, in which < 10% of patients at Month 24 had experienced TEAEs leading to discontinuation (Table 7) [43]. Interim analyses at Month 6 showed that treatment persistence was estimated to be high at 85% [29], which decreased slightly at Month 12 (75%) (Table 7) [46].

Discussion

As demonstrated in the latest data from clinical trials and real-world studies, FIL has demonstrated rapid effectiveness in the management of RA, particularly regarding clinical outcomes (lowered disease activity), symptom relief (reduced pain and fatigue), radiographic outcomes (reduced tender and swollen joints) and functional outcomes. Due to its rapid onset of action, FIL contributes to early inflammation control, which may further improve structural and clinical endpoints with longer treatment. Effectiveness is also sustained over the longer term, with RCT data extending to 8 years and real-world data demonstrating effectiveness of FIL up to Month 24.

Overall, FIL is a well-tolerated and effective treatment option for patients with RA, particularly in those with PPFs or who did not see improvements with initial MTX therapy. Nevertheless, ongoing discussion about the use of FIL in clinical practice remains, including the appropriate timing of initiation and the optimal starting dose.

Appropriate FIL Dose Depends on Patient Characteristics

To investigate current real-world practice regarding FIL dose, we can refer to treatment patterns reported in observational studies that enrolled patients who initiated any dose of FIL [25, 38, 45]. In Japan, the JCR has issued a specific guidance document on the appropriate use of FIL, stating that treatment with FIL should be initiated at 200 mg unless the patient has moderate-to-severe renal impairment (15 ≤ eGFR < 60) [11, 14]. For such patients, the JCR recommends initiating FIL at 100 mg [11, 14]. Despite this, a relatively large proportion of patients initiated FIL at 100 mg. In a Japanese real-world study, 63% of IRs to bDMARDs were prescribed a starting dose of 200 mg FIL; however, only 41% of IRs to csDMARDs were prescribed a starting dose of 200 mg FIL, while the others (59%) initiated 100 mg FIL [45]. Additionally, a post-marketing surveillance study of FIL in Japan also showed that 63.8% of patients with RA initiated FIL at 200 mg, with the remaining initiating FIL at 100 mg [47]. From a multicenter registry study in Japan, the prevalence of moderate-to-severe renal impairment (defined as eGFR < 60 ml/min/1.73 m²) among Japanese patients with RA who initiated a bDMARD or JAKi was 23.8% [48]. Based on the estimate that approximately one fourth of patients with RA may have moderate-to-severe renal impairment, initiation of FIL at 100 mg in the real world is therefore likely to be more common than recommended and not limited only to patients with moderate-to-severe renal impairment. To improve patient outcomes in RA, physicians could better adhere to guidelines for FIL dosage after careful evaluation of renal function.

While the European Summary of Product Characteristics for FIL similarly states that treatment should generally be initiated at 200 mg, there are differences compared to the JCR for recommended indications where 100 mg may be appropriate [11, 14, 41]. In Europe, older patients are recommended to initiate FIL at 100 mg, with the potential to escalate to 200 mg [41]. In a German real-world study that enrolled patients who were IRs or naïve to advanced therapies, approximately 85% of patients initiated FIL at a 200 mg dose [25]. Indeed, older age was one of the most commonly reported reasons for patients to initiate 100 mg FIL rather than 200 mg FIL, in alignment with European recommendations [25].

Generally, findings from both RCTs and real-world studies did not show statistically significant differences in clinical outcomes (e.g., disease activity measured by CDAI, SDAI, DAS28CRP and ACR20/50/70) and patient-reported outcomes (e.g., pain) by FIL dose. However, in an NMA of RCTs comparing JAKis and adalimumab in combination with MTX, 200 mg FIL + MTX was reported to be statistically superior to 100 mg FIL + MTX for ACR70 disease activity [19]. Additionally, while differences were not statistically significant, FINCH 1 demonstrated numerically better efficacy with 200 mg FIL + MTX for DAS28CRP-defined remission than adalimumab + MTX, while 100 mg FIL + MTX was similar to adalimumab + MTX [21]. Aside from clinical outcomes, studies have also reported dose-dependent reductions in radiographic progression measured by mTSS [17, 37]. Therefore, 200 mg FIL + MTX may be an optimal treatment strategy to manage structural changes in joints among patients with RA.

As the safety profile of FIL was not significantly different for 200 mg FIL compared to 100 mg FIL in the general population with RA, clinicians should opt to begin treatment with 200 mg FIL for improved radiographic outcomes. Both doses of FIL (100 mg and 200 mg) ± MTX achieved effective reductions in disease activity and radiographic progression compared with MTX monotherapy in patients with ≥ 4 PPFs and patients with eRRP [17, 18]. Incidence rates of AEs were similar between FIL doses regardless of eRRP [17]. Therefore, regardless of PPFs, patients may initiate treatment with 200 mg FIL, with the potential for dose-switching depending on their tolerance. At the time of this review, no study has been identified reporting evidence that patients with RA who achieved remission on 200 mg FIL can be maintained on a reduced 100 mg dose.

However, a potential dose-related trend was observed in older patients (≥ 65 years), primarily due to eGFR decline associated with aging [49], with reported malignancy incidence rates in RCTs numerically lower with 100 mg than with 200 mg FIL ± csDMARDs among older patients [39]. As older patients are at greater risk of TEAEs, a commonly cited reason for discontinuing FIL [25], this may also explain the persistence rates observed in real-world studies; a Japanese real-world study reported higher discontinuation rates in older patients than in those < 65 years old [45]. Based on limited real-world studies published to date [22–25, 28, 45, 46], the use of 200 mg FIL and 100 mg FIL doses in the real world appears to align with the EULAR and JCR recommendations.

Appropriate Treatment with MTX Before Starting FIL

Another consideration in real-world clinical practice is treatment with MTX before initiating FIL. In guidelines, FIL is only recommended after failing to achieve improvement with initial MTX treatment, defined as > 8 mg/week for at least 3 months by the JCR [6, 11]. Findings from the FINCH trials show that FIL + MTX reduced disease activity and radiographic progression in both IRs to MTX (i.e., those who had initial treatment with MTX) and patients who were MTX-naïve compared with MTX monotherapy [17, 18, 36]. Additionally, the proportion of patients achieving remission was similar at Week 156 regardless of MTX history [27]. Furthermore, functional outcomes, such as patient-reported HAQ-DI, were similar in both IRs to MTX and in patients who were MTX-naive in the FINCH 4 trial [27]. FIL ± MTX persistence rates were numerically higher by approximately 4% among IRs to MTX than in patients who were MTX-naïve [13]. Unfortunately, real-world data directly comparing outcomes with FIL between IRs to MTX and patients who were MTX-naive are not available. Where reported, most patients enrolled in RWE studies had prior treatment with csDMARDs [25, 45], in alignment with the recommendations from EULAR and JCR [6, 11].

As the efficacy of FIL is similar regardless of MTX history, with persistence rates among IRs similar or slightly higher than those of patients who were MTX-naive, initial therapy with MTX should remain the standard of care, with switching to FIL considered upon inadequate response or poor tolerability. However, a post-marketing surveillance study in Japan on FIL use showed that only 28.2% of patients met the JCR definition of prior MTX treatment, suggesting that some patients may not have received appropriate initial MTX treatment [47]. Physicians should thus ensure that patients with RA have received adequate initial MTX treatment (> 8 mg/week for at least 3 months) before starting FIL.

Treatment Persistence with FIL

Persistence on FIL reported in real-world studies can add value to that of RCTs as they can highlight factors impacting discontinuation that may not be considered in controlled clinical settings [25]. For example, among patients with moderate-to-severe RA in a German real-world setting, there were several reasons for discontinuing FIL, such as lack of adherence [25]. Such factors may better illuminate daily considerations for patients and their rheumatologists regarding treatment with FIL in the real world.

While findings from real-world settings largely complemented those from RCTs, there were some divergences. Notably, persistence on FIL among IRs to bDMARDs was lower in a Japanese real-world study (66%) than in the FINCH trial (85.0%) [13, 45]. This may be attributed to the broader criteria for MTX use during the look-back period for patient enrollment, under which patients were eligible for the study if they had at least one MTX prescription claim [45]. As such, patients enrolled in the Japanese real-world study may not have been adequately treated with MTX prior to initiating FIL, potentially leading to reduced response and lower persistence. However, FIL persistence among IRs to csDMARDs (including MTX) was similar between real-world patients with moderate-to-severe or difficult-to-treat RA and the FINCH 4 trial (~ 60%) [27, 45]. The divergence in persistence rates among IRs to bDMARDs may be attributed to the real-world study’s additional criteria for csDMARD history in addition to bDMARDs [45].

Overall, FIL has high persistence rates, comparable to those of other JAKis and bDMARDs [45]. This may be attributed to several frequently reported reasons for patients' preference to initiate FIL: its oral administration route and its potential to be used as monotherapy [25]. These may have led to high persistence rates among patients with FIL, as most patients with RA prefer an orally administered drug over intravenous or subcutaneous drugs [25].

Limitations

Some limitations should be noted in the interpretation of findings from this review. Findings from real-world studies may not be directly comparable to each other, or to those of the DARWIN and FINCH trials, because of differing patient and disease characteristics. For example, while the measure of outcomes such as disease activity may be defined by commonly used tools, such as CDAI, SDAI and DAS28CRP, baseline severity of RA may differ based on the definitions each study used to categorize RA (e.g., International Classification of Disease 10th revision, ACR/EULAR 2010 criteria, and differing definitions of severe or hard-to-treat RA). The real-world studies discussed in this review were generally single-country, whereas the DARWIN and FINCH trials were multi-country. Therefore, real-world findings specific to patients in a single geography may not be directly comparable to those from international RCTs.

Additionally, some limitations are inherent to the retrospective nature of the real-world studies included in this narrative review [23, 25, 28, 45]. The clinical assessments reported by retrospective observational studies were made in the context of routine clinical care [50], in contrast to endpoints assessed as part of an RCT where training may be provided to better standardize outcome assessments. Furthermore, findings from real-world studies rely on the accuracy of the data recorded in each data source, and as the data were not collected specifically for the study, issues with data accuracy or misclassification errors may be more likely during data cleaning and processing [23–25, 28, 45]. Finally, given that most of the observational studies discussed recruited patients only from a single geography, the findings may not be generalizable to the wider global patient population. FILOSOPHY is an ongoing prospective multinational observational study from Europe (Belgium, Germany, Italy, The Netherlands, Spain and the UK) [51]; while interim results have been reported, further results may be valuable to corroborate findings from single-country real-world studies. Moreover, we identified some important gaps in the evidence base, particularly for the use of FIL in real-world settings. Given that FIL was only recently approved for use in 2020 in the European Union, UK, Taiwan, South Korea, Singapore and Japan [14, 41, 52–55], follow-up time is limited [25]. Currently available real-world studies report rapid effectiveness and short-term safety, but further studies are required to understand the sustained effectiveness and safety of FIL in the real-world setting. While the prospective FILOSOPHY and PARROTFISH non-interventional studies are ongoing in Europe to assess the sustained effectiveness and safety of FIL, further studies in other geographies such as Asia are needed. Additionally, we did not identify any real-world studies measuring radiographic progression in patients initiating FIL using tools such as mTSS, which limited our discussion of FIL's real-world effectiveness in reducing radiographic progression.

Conclusion

RCTs have demonstrated the rapid and sustained effectiveness of FIL in RA for clinical, structural and functional endpoints, which are further confirmed in real-world studies. Additionally, both RCT and observational evidence have demonstrated the long-term safety of FIL in RA. Guidelines recommend 200 mg FIL + MTX following appropriate initial treatment with MTX to control disease activity and minimize joint damage. A key consideration in screening and monitoring of patients on FIL for safe management of RA in Japan may be renal impairment, especially in older patients. In Europe, risk of venous thromboembolism, major adverse cardiovascular events and malignancy have been noted as key considerations for patients initiating FIL. In the real world, additional considerations could affect outcomes with FIL, including persistence on FIL and larger variations in patient profiles and treatment backgrounds. Future results from ongoing studies can be expected to provide further insights into the long-term effectiveness, particularly on joint outcomes, and safety of FIL in the real-world context.

Author Contributions

Yoshiya Tanaka, Peter C Taylor, Jeffrey A S Ritsema, Toshihiko Kaise and Akira Kondo contributed to the conception of the review and interpretations. Yoshiya Tanaka, Peter C Taylor, Jeffrey A S Ritsema, Toshihiko Kaise and Akira Kondo reviewed and contributed to revisions in the drafting of the manuscript. Yoshiya Tanaka, Peter C Taylor, Jeffrey A S Ritsema, Toshihiko Kaise and Akira Kondo read and approved the final manuscript.

Funding

This study was sponsored by Gilead Sciences K. K., Tokyo, Japan, and Eisai Co., Ltd, Tokyo, Japan. The Journal’s Rapid Service Fee was also funded by Gilead Sciences K. K., Tokyo, Japan, and Eisai Co., Ltd, Tokyo, Japan.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declarations

Conflict of Interest

Yoshiya Tanaka has received speaking fees and/or honoraria from Chugai, UCB, Abbvie, AstraZeneca, Eli Lilly & Co, Behringer-Ingelheim, GlaxoSmithKline, Eisai, IQVIA, Daiichi-Sankyo, Otsuka, Taisho, Gilead, Bristol-Mayers. Yoshiya Tanaka is an Editorial Board member of Rheumatology and Therapy. Yoshiya Tanaka was not involved in the selection of peer reviewers for the manuscript nor any of the subsequent editorial decisions. Peter C Taylor has received research grants from Alfasigma, consulting fees from AbbVie, Aceleryn, Alfasigma, Gilead, Takeda, Janssen, Eli Lilly & Co., Nordic Pharma, Roche, and UCB, and has served on safety monitoring boards for Immunovant, Moonlake and Sanofi. Jeffrey A S Ritsema is an employee of Alfasigma Nederland BV. Toshihiko Kaise is an employee of Gilead Sciences K.K. and shareholder of Gilead Sciences, Inc. Akira Kondo is an employee of Gilead Sciences K.K. and may hold shares in Gilead Sciences, Inc.

Ethical Approval

As this review is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors, ethical approval is not required.