Anifrolumab, a monoclonal antibody to the type I interferon receptor subunit 1, for the treatment of systemic lupus erythematosus: an overview from clinical trials

Abstract

Chronic activation of the type I interferon (IFN) pathway plays a critical role in systemic lupus erythematosus (SLE) pathogenesis. Anifrolumab is a human monoclonal antibody to the type I IFN receptor subunit 1, which blocks the action of type I IFNs. Two phase 3 studies (TULIP-1 and TULIP-2) and a phase 2b study (MUSE) provide substantial evidence for the efficacy and safety of anifrolumab for moderately to severely active SLE. In all three studies, monthly intravenous anifrolumab 300 mg was associated with treatment differences >16% compared with placebo at Week 52 in British Isles Lupus Assessment Group-based Composite Lupus Assessment response rates. The combined data across a range of other clinically significant endpoints (e.g. oral corticosteroid reduction, improved skin disease, flare reduction) further support the efficacy of anifrolumab for SLE treatment. The safety profile of anifrolumab was generally similar across all studies; serious adverse events occurred in 8–16% and 16–19% of patients receiving anifrolumab and placebo, respectively. Herpes zoster incidence was greater with anifrolumab (≤7%) vs placebo (≤2%). Evidence from these clinical trials suggests that in patients with active SLE, anifrolumab is superior to placebo in achieving composite endpoints of disease activity response and oral corticosteroid reduction.

Keywords:

• Anifrolumab
• anti-interferon alpha receptor antibody
• systemic lupus erythematosus
• TULIP-1
• TULIP-2

Introduction

Systemic lupus erythematosus (SLE) is an autoimmune disease associated with substantial morbidity and mortality [1,2]. Despite a marked improvement in 10-year survival for patients with SLE over the past five decades, the all-cause mortality rate for patients with SLE is approximately 2.6-fold greater than that for the general population [3]. Among a large cohort of Japanese patients with SLE, overall 5-year survival rate was 92.2% [4]. In another smaller cohort, 5-, 10-, 15-, and 20-year survival rates were 99.3%, 94.6%, 92.0%, and 85.4%, respectively [5]. SLE severity at diagnosis is associated with mortality risk, with a hazard ratio (HR) of 3.11 (95% confidence interval [CI] 2.49, 3.89) for initially severe and 1.61 (95% CI 1.29, 2.01) for moderate SLE compared with mild SLE [6].

Despite improvements in survival rates, the management of SLE is not optimal, and complete control of disease activity is not frequently achieved [7,8], owing to the wide spectrum and severity of clinical manifestations observed in patients with SLE, association with flares (periods of heightened disease activity), and toxicity associated with long-term use of oral corticosteroids (OCS) and immunosuppressants [9,10]. Patients experience excessive risk of progressive irreversible organ damage as well as significantly reduced health-related quality of life (HRQoL), the latter of which has not improved despite advances in overall SLE prognosis [9,11]. All aspects of HRQoL are negatively affected by SLE, including physical and mental health, vitality, pain, and social and emotional functioning [10]. These reductions in HRQoL are similar to, and in some cases greater than, those associated with other chronic diseases [10,12]. Some evidence indicates that decreasing disease activity may improve HRQoL among patients with SLE [10,12,13]. Goals of SLE treatment, therefore, include improvement in physician’s assessment of disease activity, prevention of disease flares, minimization of irreversible damage accrual, avoidance of drug toxicity (including reduction of steroid use), and improvement of patient HRQoL [14].

The biologic belimumab (a B-lymphocyte stimulator-specific inhibitor) is the only agent that has received regulatory approval for the treatment of SLE in the last 60 years, despite many clinical trials of other novel treatments [15,16]. The complex pathogenesis and heterogeneous nature of the disease present challenges in clinical trial design and endpoint selection that may have contributed to trial failures and subsequent drug development failures in SLE [17,18].

Current standard-of-care (SOC) therapies for SLE include hydroxychloroquine (antimalarial), OCS, and immunosuppressants [14]. OCS have historically been the core therapy, to which other medications are added (nonsteroidal anti-inflammatory drugs [NSAIDs], topical steroids, immunosuppressants) [19,20]. Hydroxychloroquine has also been used as a mainstay for the treatment of SLE [19–21]. Belimumab, which is globally approved for the treatment of SLE, can be used in patients with inadequate response to SOC treatment (per the 2019 update to European League Against Rheumatism [EULAR] recommendations for the management of SLE) [14]. Belimumab is a human immunoglobulin (Ig) G1λ monoclonal antibody that selectively binds the soluble B-cell activating factor (BAFF; belonging to the tumor necrosis factor family), preventing binding to B-cell receptors and, in turn, preventing survival of B cells. It also reduces the differentiation of B lymphocytes into plasma cells that produce immunoglobulins [22]. Rituximab, a CD20-directed cytolytic antibody, does not have regulatory approval for the treatment of SLE but remains a treatment option for refractory disease despite negative findings from several randomized controlled trials [14,23].

SLE is characterized by dysregulation of both the innate and adaptive immune systems. This results in the breakdown of tolerance and the production of inflammatory cytokines and autoantibodies against a variety of nuclear antigens that deposit into tissues and cause injury [24,25]. Activation of the type I interferon (IFN) pathway is now recognized as playing a critical role in the pathogenesis of SLE [22]. The autoimmune complexes formed in SLE as well as toll-like receptor-mediated signaling induced by ligation with RNA and DNA secreted from neutrophil NETosis in SLE stimulate production of type I IFNs by plasmacytoid dendritic cells [22]. IFN-α (a type I IFN) promotes dendritic cell formation, activation of T cells, and production of autoantibodies by B cells [22]. Elevated levels of type I IFNs, IFN-dependent cytokines, and type I IFN-induced genes have been correlated with serologic disease activity markers (e.g. anti-double-stranded DNA antibodies, complement components) and disease flares [22,25].

Cell signaling by all type I IFNs (i.e. IFN-α, IFN-β, IFN-ε, IFN-κ, and IFN-ω) is mediated by the type I IFN-α receptor (IFNAR) [22,26]. The ongoing production of type I IFN seen in most patients with SLE is reflected by an overexpression of type I IFN-regulated genes, known as the IFN gene signature (IFNGS) [26]. Blocking the effects of type I IFNs by IFNAR blockade may reverse immune dysregulation and mitigate further tissue damage in SLE and result in symptom improvement [27].

Based on the key role of type I IFN signaling in SLE pathogenesis, IFN-α, IFNAR, and related signaling molecules have become therapeutic targets for the treatment of SLE [27]. In clinical studies, rontalizumab, a humanized IgG1 monoclonal anti-IFN-α antibody, rapidly suppressed IFN response genes, but in a phase 2 trial of patients with active SLE, it failed to meet its primary efficacy endpoint (British Isles Lupus Assessment Group [BILAG] response) [8,28]. Sifalimumab, a fully human IgG1ĸ monoclonal anti-IFN-α antibody, met the primary efficacy endpoint of SLE Responder Index [SRI(4)] response in a phase 2 trial of patients with active SLE but did not proceed into further development [8,28]. Consistent with the mixed results seen with anti-IFN-α antibodies, a vaccine preparation shown to induce neutralizing, polyclonal anti-IFN-α antibodies (IFN-Kinoid) normalized the IFNGS and reduced Lupus Low Disease Activity State in a phase 2 trial but did not significantly improve BILAG score compared with placebo [8,29,30]. Targeting the IFN receptor as opposed to IFN-α may prove a more effective approach to blocking type I IFN signaling.

Anifrolumab is a fully human IgG1κ monoclonal antibody that targets several type I IFNs through binding the type I IFN receptor subunit 1 [31]. The clinical evidence for the efficacy and safety of anifrolumab in SLE is the subject of this review. Further, investigational compounds that target downstream IFN pathways include those that inhibit Janus kinase 1 (e.g. baricitinib) or tyrosine kinase 2 (e.g. BMS-986165), both of which are activated by IFNAR [25].

Anifrolumab mode of action, pharmacodynamics, and pharmacokinetics

Anifrolumab (previously MEDI-546) is a fully human IgG1κ monoclonal antibody that blocks the action of all type I IFNs by binding to subunit 1 of the type I IFNAR (IFNAR1) with high affinity and specificity [31]. Anifrolumab also induces the internalization of IFNAR1, thereby reducing the levels of cell surface IFNAR1 available for receptor assembly. Blockade of receptor-mediated type I IFN signaling inhibits IFN-responsive gene expression and downstream inflammatory and immunological processes and events [32]. Anifrolumab administered intravenously (IV) has been studied for the treatment of adults with moderately to severely active SLE in two phase 2 studies (the MUSE randomized controlled trial and an open-label, dose-escalation study in Japanese patients) and two phase 3 studies (TULIP-1 and TULIP-2) [33–38]. The long-term safety and tolerability of anifrolumab in patients with SLE have also been evaluated in a 3-year open-label extension study of MUSE (NCT01753193).

Pharmacodynamics

In the MUSE trial, among the 70.5% of patients who were positive for IFNGS at baseline, expression of IFNGS in whole blood decreased after anifrolumab administration for all dosages [39]. The median suppression of a 21-gene type I IFN pharmacodynamic signature at Week 24 was 89.7% and 91.7% in the 300-mg and 1000-mg every 4 weeks (Q4W) dosage arms, respectively. Analysis of blood samples from patients with SLE enrolled in the MUSE trial also indicated that anifrolumab treatment lowered concentrations of IFN-induced chemokines associated with B, T, and other immune cell migration, and of proteins associated with endothelial activation that were dysregulated at baseline [40]. Furthermore, anifrolumab treatment was associated with a reversal of SLE-associated lymphopenia, neutropenia, monocytopenia, and thrombocytopenia [40]. In the two phase 3 trials (TULIP-1 and TULIP-2), suppression of the IFNGS was achieved early in patients with high IFNGS at baseline who received anifrolumab (median suppression of a 21-gene type I IFN pharmacodynamic signature at Week 12 was >80%) and was maintained through Week 52 [36,37]. There was a trend toward normalization of serologic changes from baseline in double-stranded DNA antibodies and complement C3, C4, and CH50 in TULIP-1 and TULIP-2 (Table 1) [36,37].

Table 1. Results for serologic changes from baseline at week 52 in TULIP-1 and TULIP-2 [36,37].

	TULIP-1		TULIP-2
Biomarker	Placebo	Anifrolumab 300 mg	Placebo	Anifrolumab 300 mg
Percentage of baseline IFN 21-gene signature in IFNGS test-high patients at baseline, median (range) [n]	89.178 (62.000, 129.086) [114]	11.771 (5.568, 26.683) [97]	95.441 (66.63, 124.78) [112]	8.906 (5.84, 23.05) [117]
Change from baseline in anti-dsDNA antibodies in patients with elevated anti-dsDNA at baseline, median U/mL (range) [n]	−6.6 (−18.0, 19.4) [62]	−19.6 (−53.5, −7.3) [62]	−4.650 (−32.84, 2.60) [49]	−12.560 (−24.87, −0.41) [71]
Change from baseline in complement concentration in patients with abnormal complement levels at baseline
C3, median g/L (range) [n]	0.0410 (−0.0750, 0.1020) [49]	0.1305 (0.0510, 0.2140) [48]	0.0475 (−0.0500, 0.117) [58]	0.0890 (−0.0160, 0.2130) [64]
C4, median g/L (range) [n]	0.0125 (0.0000, 0.0330) [30]	0.0190 (0.0025, 0.0310) [28]	0.0060 (−0.0020, 0.0170) [33]	0.0100 (0.0010, 0.0370) [43]
CH50, median CH50 units (range) [n]	89.0 (40.5, 162.5) [8]	112.5 (43.0, 197.0) [12]	75.0 (40, 100) [13]	106.0 (12, 143) [14]

C3: complement component 3; C4: complement component 4; CH50: total complement levels; dsDNA: double-stranded DNA; IFN: interferon; IFNGS: interferon gene signature.

Pharmacokinetics

In the MUSE trial (N = 305), anifrolumab exhibited greater than dose-proportional pharmacokinetic exposure when comparing the dosages of 300 mg IV Q4W and 1000 mg IV Q4W owing to target-mediated clearance of the drug-target complex [39]. On Day 169, mean ± standard deviation C_trough (lowest anifrolumab concentration between dosing) was 18.8 ± 10.9 μg/mL and 115 ± 62.9 μg/mL, respectively, for the 300-mg Q4W and 1000-mg Q4W dosages [39]. In the open-label, dose-escalation study in Japanese patients (N = 20), anifrolumab demonstrated nonlinear pharmacokinetics after the first and last dose and suppressed the IFNGS in a dose-dependent manner [38]. This trial consisted of two stages: Stage 1, treatment and follow-up, and Stage 2, long-term treatment. In Stage 1, patients received a single intravenous dose (100, 300, or 1000 mg) of anifrolumab at Day 1 and were followed up at Day 29. Patients then received repeated Q4W dosing from Days 29 to 337 and were subsequently followed up for 85 days (Days 337–422). In Stage 1, first dose maximum concentration (C_max) increased from 42.4 µg/mL with the 100-mg dosage to 259.2 µg/mL with the 1000-mg dosage, and mean area under the curve (AUC) during the dose interval increased from 152.3 µg⋅day/mL with 100 mg to 5814.6 µg⋅day/mL with 1000 mg. At the last dose, the maximum concentration under these conditions (C_{ss, max}) was 25.0 µg/mL and 414.3 µg/mL for 100 mg and 1000 mg, respectively, and AUC at the end of dosing (AUC_τ) was 152.3 µg⋅day/mL and 5814.6 µg⋅day/mL for 100 mg and 1000 mg, respectively.

Anifrolumab clinical efficacy

Two phase 3 studies (TULIP-1 and TULIP-2) and one phase 2b study (MUSE), in which evaluation of efficacy was the primary study objective, provide cumulative evidence in a large patient group for the efficacy of anifrolumab via intravenous infusion for moderately to severely active SLE [35–37]. All three studies were global, randomized, double-blind, and placebo controlled, with highly similar study designs, including a 52-week treatment period, and the same patient population. The studies were conducted with consistent inclusion and exclusion criteria to identify patients who had moderately to severely active SLE (measured by SLE Disease Activity Index 2000 [SLEDAI-2K] score ≥6, plus BILAG-2004 ≥ 1 A or ≥2 B organ domain scores, plus Physician’s Global Assessment [PGA] score ≥1), despite having received acceptable and stable dosages of SOC SLE treatment. Patients with active severe lupus nephritis or neuropsychiatric SLE were excluded in all studies. Additional key exclusion criteria were history of a primary immunodeficiency, splenectomy, or underlying condition predisposing to infection; positive result for human immunodeficiency virus infection; positive test for hepatitis B or C; severe herpes infection any time before Day 1; and any recent clinically significant infection. Patients needed to satisfy the American College of Rheumatology classification criteria for SLE at screening, including a positive test for antinuclear antibodies, anti-double-stranded DNA antibodies, or anti-Smith antibodies. All patients were required to receive SOC treatment with at least one of the following: an OCS, antimalarial, immunosuppressant, or a combination thereof, in addition to anifrolumab or placebo during the treatment period. A type I IFNGS test (validated 4-gene test) was used during screening to identify the subgroup of patients with persistent overexpression of the type I IFNGS (IFNGS test-high patients).

Baseline demographics and disease characteristics of the patients enrolled were broadly similar across all three studies and were representative of the target population of patients with moderately to severely active SLE. The mean age ranged from 39.1 to 43.1 years, and >90% of patients were female. Patients had a mean Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) global score ranging from 10.7 to 11.5 and in TULIP-1 and TULIP-2, 69% to 73% of patients were reported to have a score of ≥10%. In TULIP-1 and TULIP-2, ≥45% of patients had severe disease in ≥1 organ system (BILAG-2004 A score) and ≥43% had moderate disease in ≥2 organ systems (BILAG-2004 B score). In MUSE, patients’ mean BILAG 2004 global score ranged from 19.6 to 19.8. The majority of patients across the studies had active skin disease (mean Cutaneous Lupus Erythematosus Disease Area and Severity Index [CLASI] score 6.7–8.5) and had ≥6.2 swollen joints and ≥9.0 tender joints. Most patients were also classified as type I IFNGS test-high (74.5–83.3%). Between 66.1% and 82% of patients were receiving antimalarials and the mean OCS dosage at baseline ranged from 10.6 mg/day to 12.8 mg/day. The United States, Europe, Latin America, and Asia-Pacific regions are well represented in the anifrolumab SLE clinical program; in TULIP-1 and TULIP-2 approximately one-third of patients in were from North America.

The primary objective in all three studies was to evaluate overall disease activity. Stratification factors were used in all three studies to ensure balance among treatment groups at baseline, including disease severity (SLEDAI-2K <10 points vs ≥10 points), OCS dosage on Day 1 (prednisone <10 mg/day vs ≥10 mg/day or equivalent), and type I IFNGS test status at screening (high vs low).

MUSE (NCT01438489) was a phase 2b, multicenter, double-blind, randomized, placebo-controlled, 52-week study of anifrolumab 300 mg and 1000 mg compared with placebo in adult patients with moderately to severely active SLE despite SOC therapy [35]. OCS tapering was encouraged during the trial, but was at the discretion of the investigators. The primary efficacy endpoint was evaluated based on reduction in SLE disease activity measured by SRI(4) with sustained reduction in OCS use after 24 weeks of treatment [35].

The pivotal TULIP (Treatment of Uncontrolled Lupus via the IFN Pathway) program comprised two phase 3, multinational, randomized, double-blind, placebo-controlled, parallel-group clinical trials, TULIP-1 and TULIP-2 [36,37]. TULIP-1 (NCT02446912) compared anifrolumab 150 mg and 300 mg with placebo in adult patients with moderately to severely active SLE despite SOC treatment [36]. Efficacy was evaluated based on reduction in SLE disease activity as measured by SRI(4) response. TULIP-2 (NCT02446899) compared anifrolumab 300 mg with placebo in adult patients with moderately to severely active SLE despite SOC treatment [37]. Efficacy in this trial was evaluated based on reduction in SLE disease activity as measured by BILAG-based Composite Lupus Assessment (BICLA) response.

TULIP-1 and TULIP-2 study designs were nearly identical (Figure 1). Both studies consisted of a 52-week treatment period with anifrolumab or placebo IV Q4W from Week 0 to Week 48 for a total of 13 doses. Primary endpoint was evaluated at Week 52. OCS tapering attempts were mandated in both studies between Weeks 8 and 40 for patients receiving baseline oral prednisone ≥10 mg/day or equivalent, until a dosage ≤7.5 mg/day was achieved, which had to be sustained through to Week 52 [36,37].

Figure 1. TULIP-1 and TULIP-2 study design. ACR: American College of Rheumatology; ANA: antinuclear antibodies; anti-dsDNA: anti-double-stranded DNA; anti-Sm: anti-Smith antibodies; BICLA: BILAG-based Composite Lupus Assessment; BILAG: British Isles Lupus Assessment Group; IFNGS: interferon gene signature; IV: intravenous; OCS: oral corticosteroid; PGA: Physician’s Global Assessment; Q4W: every 4 weeks; SLE: systemic lupus erythematosus; SLEDAI-2K: SLE Disease Activity Index 2000; SRI(4): SLE Responder Index. ^aEligible patients fulfilled ACR classification for SLE; ^bPatients were stratified by IFNGS status, SLEDAI-2K score, and OCS dosage; ^cFor patients with baseline OCS of prednisone ≥10 mg/day or equivalent.

Composite endpoints that detected clinically meaningful improvement of SLE disease activity were used in both studies: SRI(4) (primary endpoint in TULIP-1) and BICLA (primary endpoint in TULIP-2). In TULIP-2, a protocol amendment changed the primary endpoint from SRI(4) to BICLA response before unblinding of trial data and after completion of TULIP-1 [37]. This change was informed by MUSE and TULIP-1 analyses. During the design of both phase 3 trials, SRI(4) and BICLA response were both candidates for primary endpoints; SRI(4) was selected because the SRI was the primary endpoint in MUSE and phase 3 trials of other targeted therapies (including belimumab) [41]. Data from the TULIP-2 trial were not used to inform the decision to alter the primary or secondary endpoints. The timing of the change in trial endpoints is detailed in Figure 2.

Figure 2. Key activities leading to changes to the primary endpoint and key secondary endpoints in the phase 3 TULIP-2 trial. BICLA: BILAG-based Composite Lupus Assessment; BILAG: British Isles Lupus Assessment Group; SRI(4): SLE Responder Index.

In addition, after a review of TULIP-1, which did not meet its primary endpoint, and before unblinding of TULIP-2, it became apparent that the protocol-restricted medication rules, which were a component of the composite efficacy endpoints, had misclassified some patients as nonresponders [36]. An error was made and some patients who had used a new NSAID or had an increased NSAID dosage were misclassified. Unlike immunosuppressants and antimalarials, NSAIDs are not disease-modifying agents; they are used for temporary symptom relief. It was, therefore, considered inappropriate to have classified patients as nonresponders if they used new or increased NSAID dosages, and this finding led to a review of all restricted medication rules to ensure that they were clinically appropriate. Notably, restricted medication rules (i.e. no use of restricted medication beyond protocol-allowed thresholds) are a component of several composite efficacy endpoints (e.g. SRI, BICLA, OCS reduction, improvements in CLASI score) with potentially substantial effects on outcome interpretation. Based on the results of the review, revised restricted medication rules for the use of NSAIDs and other medications were agreed on with an external steering committee for the TULIP program; the statistical analysis plan for TULIP-2 was updated accordingly before study unblinding, and TULIP-1 key analyses were repeated (post hoc) using the amended restricted medication rules. Results were reported in the publication reporting TULIP-1 for both analyses, and the primary endpoint of SRI(4) response was not significantly affected by use of the amended restriction rules [34].

MUSE phase 2b trial efficacy results

The MUSE trial was conducted to evaluate the efficacy and safety of two fixed intravenous dosages of anifrolumab in adults with moderately to severely active SLE with inadequate responses to SOC therapy [35]. Patients (N = 305) were randomized to receive anifrolumab (300 mg or 1000 mg) or placebo IV, in addition to SOC therapy, Q4W for 48 weeks. The primary endpoint was a composite of SRI(4) response at Week 24 with sustained OCS reduction (<10 mg/day and less than or equal to the dosage at Week 1 from Weeks 12 through 24). SRI(4) response was defined as a ≥ 4-point reduction in SLEDAI-2K score, no new BILAG A organ domain score or >1 new BILAG B organ domain score, and <0.3-point increase in PGA score from baseline. Secondary endpoints assessed at Week 52 included SRI(4) response rates in patients with sustained OCS reduction and reduction of OCS dosage at Week 52 to ≤7.5 mg/day in patients receiving ≥10 mg/day at baseline. Other secondary efficacy measures assessed at Week 52 included percentage of patients with ≥50% improvement in CLASI score, BICLA response, ≥50% improvement in swollen and tender joint counts, and modified SRI requiring SLEDAI-2K reductions of 5–8 points to be considered a responder [35].

The primary endpoint of a composite of SRI(4) response with sustained OCS reduction at Week 24 was met by more patients receiving anifrolumab (34.3% and 28.8% for patients receiving 300 mg [n = 99] and 1000 mg [n = 104], respectively) than placebo (17.6% [n = 102]; p = .014 and p = .063 for 300 mg and 1000 mg vs placebo, respectively). A greater effect size was observed in patients with a high IFNGS at baseline, with 36.0% (p = .004) and 28.2% (p = .029) of patients treated with anifrolumab 300 mg and 1000 mg, respectively, achieving the primary endpoint vs 13.2% of patients receiving placebo. In patients with a low IFNGS at baseline, the respective response rates for patients achieving the primary endpoint were 29.2%, 30.8%, and 30.8% for anifrolumab 300 mg, anifrolumab 1000 mg and placebo [35].

The secondary outcome of SRI(4) response at Week 52 with sustained OCS reduction was achieved by 51.5% of patients receiving anifrolumab 300 mg and 38.5% receiving anifrolumab 1000 mg, compared with 25.5% of patients receiving placebo (p < .001 and p = .048 for 300 mg and 1000 mg vs placebo, respectively) [35]. A greater treatment effect for anifrolumab compared with placebo was observed in patients with a high vs low IFNGS at baseline, with 52.0% (p < .001) and 38.5% (p = .013) of IFNGS test-high patients treated with anifrolumab 300 mg and 1000 mg, respectively, achieving SRI(4) response vs 19.7% of patients receiving placebo. In IFNGS test-low patients at baseline, 50.0% (p = .514) and 38.5% (p = .849) of patients treated with anifrolumab 300 mg and 1000 mg, respectively, achieved SRI(4) response vs 42.3% of patients receiving placebo [35]. A greater percentage of patients receiving anifrolumab 300 mg and 1000 mg also achieved BICLA response at Week 52 compared with those receiving placebo (53.5%, 41.2%, and 25.7%, respectively; p < .001 for 300 mg and p = .018 for 1000 mg vs placebo) (Figure 3) [35]. A greater effect size for BICLA response was seen in patients with high IFNGS at baseline with anifrolumab 300 mg vs placebo (odds ratio [OR] 3.65; 95% CI 2.02, 6.60; p < .001) and 1000 mg (OR 2.41; 95% CI 1.34, 4.35; p = .014) compared with patients with low IFNGS at baseline for anifrolumab 300 mg (OR 3.19; 95% CI 1.16, 8.73; p = .059) and 1000 mg (OR 1.38; 95% CI 0.51, 3.70; p = .596) [35]. Sustained OCS reduction, reduction in severity of skin disease (reduction in CLASI score), and improvement in tender and swollen joint counts also favored anifrolumab and were statistically significantly greater with anifrolumab 300 mg vs placebo (Figure 3) [35]. Among patients receiving prednisone ≥10 mg/day or equivalent at baseline, 56.4% and 31.7% of those receiving anifrolumab 300 mg and 1000 mg, respectively, had reduced OCS dosages to ≤7.5 mg/day at Week 52 compared with 26.6% receiving placebo (p = .001 for 300 mg vs placebo; nonsignificant for 1000 mg) [35]. Among patients with at least moderately active skin disease (CLASI score ≥10) at baseline, a ≥ 50% reduction in CLASI score at Week 12 occurred in 63.0% and 58.3% of those receiving anifrolumab 300 mg and 1000 mg, respectively, compared with 30.8% receiving placebo (p = .013 for 300 mg vs placebo; nonsignificant for 1000 mg) [35]. A ≥50% improvement in joint counts was experienced by 69.6%, 64.6%, and 48.6% of patients receiving anifrolumab 300 mg, anifrolumab 1000 mg, and placebo, respectively, among those with ≥8 swollen and ≥8 tender joints at baseline (p = .038 for 300 mg vs placebo; nonsignificant for 1000 mg) [35].

Figure 3. Overall efficacy results for TULIP-1, TULIP-2, and MUSE [35–37]. BICLA: BILAG-based Composite Lupus Assessment; BILAG: British Isles Lupus Assessment Group; CI: confidence interval; CLASI: Cutaneous Lupus Erythematosus Disease Area and Severity Index; IFNGS: interferon gene signature; OCS: oral corticosteroid; SRI(4): SLE Responder Index. Analytic methods and definitions differ across trials. ^aPublished data expressed as odds ratio; ^bPreviously unpublished data.

TULIP-1 phase 3 trial

In TULIP-1, 457 patients were randomly assigned (2:1:2) to receive placebo, anifrolumab 150 mg, or anifrolumab 300 mg IV (300 mg being the chosen therapeutic dosage and 150 mg included to provide dose–response data) in addition to SOC treatment [36]. The primary efficacy outcome was the proportion of patients who achieved an SRI(4) response at Week 52 in the anifrolumab 300-mg vs placebo groups (Table 2). SRI(4) response was defined as ≥4-point reduction in SLEDAI-2K, less than 1 new BILAG-2004 A or less than 2 new BILAG-2004 B organ domain scores, <0.3-point increase in PGA score from baseline, no use of restricted medications beyond protocol-allowed thresholds, and no discontinuation of investigational product [36].

Table 2. Primary and key secondary endpoints in TULIP-1 and TULIP-2.

	TULIP-1	TULIP-2
Primary endpoint	SRI(4) response at Week 52	BICLA response at Week 52^a
Key secondary endpoints	SRI(4) response in IFNGS test-high patients at Week 52	BICLA response in IFNGS test-high patients at Week 52
	Sustained OCS tapering at Week 52	Sustained OCS tapering at Week 52
	CLASI response at Week 12	CLASI response at Week 12
	SRI(4) response at Week 24	Improvement in joints at Week 52^a
	Annualized flare rate	Annualized flare rate

BICLA: BILAG-based Composite Lupus Assessment; BILAG: British Isles Lupus Assessment Group; CLASI: Cutaneous Lupus Erythematosus Disease Area and Severity Index; IFNGS: interferon gene signature; OCS: oral corticosteroid; SLE: systemic lupus erythematosus; SRI(4): SLE Responder Index.

^aA prespecified, secondary endpoint in TULIP-1, not adjusted for multiplicity.

Key secondary endpoints adjusted for multiplicity included the proportion of patients achieving SRI(4) response at Week 52 in the IFNGS test-high subgroup, sustained OCS dosage reduction to ≤7.5 mg/day from Weeks 40 to 52 in patients receiving OCS ≥10 mg/day at baseline, ≥50% reduction in CLASI score by Week 12 in patients with a CLASI activity score ≥10 at baseline, and SRI(4) response at Week 24, as well as annualized flare rate through Week 52. Among prespecified secondary outcomes not adjusted for multiplicity were the proportion of patients achieving BICLA response, proportion of patients achieving SRI(5−8) responses, ≥50% reduction in swollen and tender joint counts, mean change from baseline in PGA score, and mean change from baseline in SLEDAI-2K total score. A post hoc analysis exploring the time to first BICLA response sustained through Week 52 was performed using a Cox proportional hazard model [36].

In the analysis with amended rules for restricted medications, the proportion of patients at Week 52 achieving the primary outcome of SRI(4) response was comparable between anifrolumab 300-mg (84/180 [47%]) and placebo groups (79/184 [43%]; difference −3.9; 95% CI −6.3, 14.1; p = .45) (Figure 3). Similarly, in the prespecified analysis (unamended restricted medication rules), the proportion of patients at Week 52 with an SRI(4) response was 65 (36%) of 180 treated with anifrolumab 300 mg and 74 (40%) of 184 in the placebo group (difference −4.2; 95% CI −14.2, 5.8; p = .41) [36].

Because the primary endpoint was not met, secondary endpoints were not formally tested per the statistical analysis plan. However, several secondary outcomes were achieved in a numerically greater proportion of patients receiving anifrolumab 300 mg vs placebo in the analysis with amended restricted medication rules (Figure 3) [36]. SRI(4) response at Week 52 in the IFNGS test-high subgroup was achieved by 48% (71/148) of patients receiving anifrolumab and 42% (63/151) receiving placebo (adjusted difference 6.4; 95% CI −4.8, 17.7; nominal p = .261). Sustained OCS reduction to ≤7.5 mg/day was achieved by 49% (50/103) of patients receiving anifrolumab and 32% (33/102) receiving placebo (adjusted difference 16.7%; 95% CI 3.5, 29.8; nominal p = .013) among those receiving prednisone ≥10 mg/day or equivalent at baseline. A ≥ 50% reduction in CLASI score at Week 12 occurred in 43% (25/58) of patients receiving anifrolumab and 26% (14/54) receiving placebo (adjusted difference 18.7%; 95% CI 1.4, 36.0; nominal p = .034) among those with at least moderately active skin disease (CLASI score ≥10) at baseline. Among patients with ≥8 swollen and ≥8 tender joints at baseline, 53% (37/70) in the anifrolumab group and 32% (22/68) in the placebo group (adjusted difference 20.7%; 95% CI 4.7, 36.7) had a ≥ 50% reduction in both swollen and tender joint counts at Week 52. BICLA response at Week 52 was achieved by more patients in the anifrolumab 300-mg group (83/180; 46%) than in the placebo group (54/184; 30%; adjusted difference 16.4%; 95% CI 6.7, 26.2). The HR for time to attainment of sustained BICLA responses up to Week 52 favored the anifrolumab 300-mg group over placebo (HR 1.93; 95% CI 1.38, 2.73). In addition, the BILAG-based annualized flare rate was numerically lower for anifrolumab (0.60) than placebo (0.72; rate ratio 0.83; 95% CI 0.60, 1.14; nominal p = .258) [36]. Therefore, although the primary endpoint was not reached in this study, several clinically important secondary endpoints, including OCS dosage reduction, CLASI response, and BICLA response (a rigorous composite global disease measure), suggested clinical benefit of anifrolumab compared with placebo.

TULIP-2 phase 3 trial

In TULIP-2, 362 patients were randomly assigned 1:1 to receive anifrolumab 300 mg or placebo IV in addition to SOC treatment [37]. The primary efficacy endpoint in TULIP-2 was the difference between treatment groups in percentage of patients who had a BICLA response at Week 52, defined as all of the following: reduction of all severe (BILAG-2004 A) or moderately severe (BILAG-2004 B) disease activity at baseline to lower levels (BILAG-2004 B, C, or D, and C or D, respectively) and no worsening in other organ systems (worsening defined as ≥1 new BILAG-2004 A item or ≥2 new BILAG-2004 B organ domain scores); no worsening in disease activity, as determined by SLEDAI-2K score (no increase from baseline) and PGA score (no increase ≥0.3 points from baseline); no discontinuation of the trial intervention; and no use of restricted medications beyond protocol-allowed thresholds (Table 2) [37].

Key secondary endpoints adjusted for multiplicity included achievement of BICLA response at Week 52 in patients with a high IFNGS at baseline; OCS dosage reduction to ≤7.5 mg/day from Weeks 40 to 52 in patients receiving OCS ≥10 mg/day at baseline; ≥50% reduction in CLASI score by Week 12 in patients with a CLASI activity score ≥10 at baseline; ≥50% reduction in swollen and tender joint counts at Week 52; and annualized flare rate through Week 52 (Table 2). The time to onset of a BICLA response that was sustained through Week 52 was evaluated using a Cox proportional hazard model and not adjusted for multiple comparisons [37].

A higher percentage of patients achieved the primary outcome of BICLA response at Week 52 in the anifrolumab group (47.8%) than in the placebo group (31.5%; adjusted difference 16.3%; 95% CI 6.3, 26.3; p = .001) (Figure 3). In the IFNGS test-high subgroup, the percentage of patients with BICLA response at Week 52 was 48.0% (72/150) in the anifrolumab group and 30.7% (46/151) in the placebo group (adjusted difference 17.3%; 95% CI 6.5, 28.2; adjusted p = .002) [37]. The corresponding results in the IFNGS test-low subgroup were 46.7% (14/30) and 35.5% (11/31) of patients in the anifrolumab and placebo groups, respectively (adjusted difference 11.2; 95% CI −13.5, 35.8) [37]. There was uniformity of BICLA responses favoring anifrolumab across the other protocol-defined patient subgroups of baseline disease severity, race, ethnicity, age, sex, age at disease onset, and anti-drug antibody status; however, the study was not powered to show a difference in these subgroups and sample sizes were small in some instances [37]. The HR for time to attainment of sustained BICLA responses up to Week 52 in the overall group favored the anifrolumab 300-mg group over placebo (HR 1.55; 95% CI 1.11, 2.18) [37].

Anifrolumab also showed a significant benefit for sustained OCS reduction and reduction in severity of skin disease (reduction in CLASI score) (Figure 3) [37]. Among patients receiving prednisone ≥10 mg/day or equivalent at baseline, 51.7% (45/87) of anifrolumab-treated patients and 30.1% (25/83) of those receiving placebo achieved a sustained reduction to ≤7.5 mg/day (adjusted difference 21.2%; 95% CI 6.8, 35.7; adjusted p = .01). Among patients with at least moderately active skin disease (CLASI score ≥10) at baseline, 49.0% (24/49) of those receiving anifrolumab and 25.0% (10/40) of those receiving placebo experienced CLASI score reductions ≥50% at Week 12 (adjusted difference 24.0%; 95% CI 4.3, 43.6; adjusted p = .04) [37]. Although treatment response in organs other than skin and joints was not part of the prespecified analyses, the BICLA response definition requires improvement in all organ systems affected at baseline (reduction of all baseline BILAG-2004 A and B domain scores to B/C/D and C/D, respectively) and no new flares in the remaining BILAG-2004 organ systems. At baseline, the most affected organ domains for enrolled patients were mucocutaneous and musculoskeletal (>80% with BILAG-2004 A or B involvement). Baseline BILAG-2004 A or B scores occurred less frequently in the cardiorespiratory, constitutional, renal, neuropsychiatric, gastrointestinal, hematologic, and ophthalmic domains. Therefore, by definition, response occurred across all these affected BILAG-2004 organ systems in patients who achieved BICLA response. The BILAG-2004–based annualized flare rate was 0.43 in the anifrolumab group and 0.64 in the placebo group (adjusted rate ratio 0.67; 95% CI 0.48, 0.94; adjusted p = .08) [37]. Among patients with ≥6 swollen and ≥6 tender joints at baseline, 42.2% (30/71) in the anifrolumab group and 37.5% (34/90) in the placebo group (adjusted difference 4.7%; 95% CI −10.6, 20.0; adjusted p = .55) had a ≥ 50% reduction in both swollen and tender joint counts at Week 52 [37].

Summary of evidence for anifrolumab clinical efficacy in SLE

TULIP-2 provides strong evidence for anifrolumab efficacy for the treatment of patients with moderately to severely active SLE despite SOC treatment based on the prespecified BICLA primary endpoint. The combined data from all three clinical trials further support the efficacy of anifrolumab 300 mg in these patients across a range of clinically significant endpoints. Importantly, there was a consistent benefit of anifrolumab 300 mg in all studies for BICLA response rate at Week 52; treatment differences >16% compared with placebo were observed in BICLA response rates in all three studies. In addition, in both TULIP-2 and MUSE, anifrolumab suggested treatment benefit for SRI(4) response. The consistent, supporting efficacy evidence for anifrolumab 300 mg across the many key secondary endpoints (e.g. OCS reduction, improvement in CLASI score, flare reduction) is relevant to characterize the full extent of anifrolumab’s treatment effect, particularly given the heterogeneity of SLE manifestations. Anifrolumab shows efficacy for the reduction of flares, and the onset of treatment effect for reducing disease activity occurs as early as 8–12 weeks after treatment initiation, when numerical separation of BICLA response rates by >10% were observed in favor of anifrolumab 300 mg and remained throughout 52 weeks of treatment. In addition, the steroid-sparing effect of anifrolumab potentially reduces the cumulative risk of long-term organ damage associated with SLE [42]. The improvements seen with anifrolumab treatment in skin manifestations (CLASI activity score) are also particularly important as they are common, and the face, head, and neck are frequently involved with lesions that are visible.

Anifrolumab safety and tolerability

Safety and tolerability of anifrolumab were consistent and generally similar across all three efficacy studies (Table 3) [35–37]. The percentage of patients with any adverse event (AE) ranged from 85% to 89% across the studies among those treated with anifrolumab, and from 77% to 84% in placebo groups. The most common AEs included upper respiratory tract infections, nasopharyngitis, and infusion-related reactions. Anifrolumab infusions were generally well tolerated, and there was one report of anaphylaxis that occurred in a patient receiving anifrolumab 150 mg in TULIP-1. Few patients had hypersensitivity, and most infusion-related AEs were not serious and were mild or moderate in intensity. Serious AEs (SAEs) occurred in 8–16% of anifrolumab-treated patients and in 16–19% of patients receiving placebo. There was one death each in the treatment period of TULIP-1 and TULIP-2, both of which occurred in the anifrolumab treatment arm and were due to pneumonia. There was also one death in MUSE of a patient who received 1 dose of anifrolumab 1000 mg and had acute colitis. The percentage of patients with AEs leading to discontinuation was smaller among those receiving anifrolumab vs placebo in TULIP-2 and MUSE but larger among anifrolumab-treated patients in TULIP-1 (6% vs 3% with placebo).

Table 3. Adverse events during treatment period in TULIP-1, TULIP-2, and MUSE [35–37].

	TULIP-1		TULIP-2		MUSE
Event, n (%)	Placebo (n = 184)	Anifrolumab 300 mg (n = 180)	Placebo (n = 182)	Anifrolumab 300 mg (n = 180)	Placebo (n = 101)	Anifrolumab 300 mg (n = 99)
Any AE	144 (78)	161 (89)	153 (84)	159 (88)	78 (77)	84 (85)
Serious AE	30 (16)	25 (14)	31 (17)	15 (8)	19 (19)	16 (16)
Death	0	1 (1)^a	0	1 (1)^a	0	0
AE leading to discontinuation of intervention	5 (3)	11 (6)	13 (7)	5 (3)	8 (8)	3 (3)
AE of special interest	–	–	18 (10)	25 (14)	12 (12)	10 (10)
Herpes zoster	3 (2)	10 (6)	2 (1)	13 (7)	2 (2)	5 (5)
Nonopportunistic serious infections	8 (4)	9 (5)	10 (6)	5 (3)	N/A	N/A
Opportunistic infection	1 (1)	1 (1)	0	0	N/A	N/A
Influenza	2 (1)	2 (1)	6 (3)	4 (2)	2 (2)	6 (6)
Tuberculosis	1 (1)	1 (1)	0	3 (2)	N/A	N/A
Anaphylaxis	0	0	0	0	N/A	N/A
Major adverse cardiovascular event	0	0	0	1 (1)	N/A	N/A
Vasculitis	0	0	0	0	N/A	N/A
Cancer^b	1 (1)	3 (2)	1 (<1)	0	N/A	N/A
Serious AE occurring in ≥2 patients in the trial
SLE worsening^c	3 (2)	3 (2)	6 (3)	1 (1)	N/A	N/A
Pneumonia	1 (1)	3 (2)	7 (4)	3 (2)	N/A	N/A
Urinary tract infection	1 (1)	1 (1)	NR	NR	N/A	N/A
Asthma	0	2 (1)	NR	NR	N/A	N/A
Chest pain	0	2 (1)	NR	NR	N/A	N/A
Gastroenteritis, viral	NR	NR	0	2 (1)	N/A	N/A
Radius fracture	NR	NR	2 (1)	0	N/A	N/A
AE with frequency >5% in the anifrolumab group
Upper respiratory tract infection	18 (10)	22 (12)	18 (10)	39 (22)	10 (10)	13 (13)
Nasopharyngitis	22 (12)	36 (20)	20 (11)	28 (16)	4 (4)	12 (12)
Infusion-related reaction	13 (7)	16 (9)	14 (8)	25 (14)	6 (6)	2 (2)
Bronchitis	10 (5)	16 (9)	7 (4)	22 (12)	4 (4)	7 (7)
Urinary tract infection	27 (15)	22 (12)	25 (14)	20 (11)	11 (11)	15 (15)
Herpes zoster	3 (2)	10 (6)	2 (1)	13 (7)	2 (2)	5 (5)^d
Sinusitis	NR	NR	9 (5)	12 (7)	3 (3)	6 (6)
Arthralgia	NR	NR	6 (3)	10 (6)	NR	NR
Back pain	NR	NR	3 (2)	10 (6)	NR	NR
Cough	NR	NR	6 (3)	10 (6)	NR	NR
Headache	16 (9)	17 (9)	NR	NR	13 (13)	12 (12)
Pharyngitis	13 (7)	12 (7)	NR	NR	NR	NR
Hypersensitivity	2 (1)	11 (6)	1 (1)	2 (1)	NR	NR
Influenza	NR	NR	NR	NR	2 (2)	6 (6)

AE: adverse event; N/A: not available; NR: not reported at this frequency; SLE: systemic lupus erythematosus.

^aTULIP-1: Death due to pneumonia; patient received 2 doses of anifrolumab 300 mg. TULIP-2: Death due to pneumonia.

^bMalignancies were squamous cell carcinoma of the cervix in the placebo group and squamous cell carcinoma, squamous cell carcinoma of skin, and B-cell lymphoma in the anifrolumab 300-mg group.

^cWorsening of SLE was captured in disease activity indexes. It was captured as an AE only if it also met the definition of a serious AE (i.e. it resulted in death, was immediately life-threatening, required inpatient hospitalization or prolonged existing hospitalization, resulted in persistent or clinically significant disability or incapacity, or was an important medical event that may jeopardize the patient or may require medical intervention to prevent one of the outcomes listed above). Inclusion of worsening of SLE as an AE biases results.

^dOne patient also had transverse myelitis with a quantitatively positive test result for varicella zoster virus in the cerebrospinal fluid.

There was an increased incidence of herpes zoster in the anifrolumab treatment groups (5–7%) compared with placebo groups (1–2%); most occurrences were cutaneous, not serious, and did not lead to discontinuation. All responded to SOC treatment and generally resolved without sequelae. For other AEs of special interest, the incidence was low and similar across treatment groups.

Of 20 patients enrolled in Stage 1 of the open-label, dose-escalation study in Japanese patients, 17 received anifrolumab 100 mg (n = 6), 300 mg (n = 5), or 1000 mg (n = 6) IV. AE and SAE incidences were similar between dosage cohorts. SAEs occurred in 7/17 (Stage 1) and 3/9 (Stage 2) patients; AEs leading to discontinuation occurred in 4/17 (Stage 1) and 2/9 (Stage 2) patients [38].

Expert opinion

The goals of SLE management include remission (or at least low disease activity), prevention of disease flares, minimization of organ damage accrual, and reduction of adverse effects associated with treatment (especially OCS) [25]. Ideal treatment, therefore, improves disease activity, allows for steroid reduction, and minimizes flares. Unmet needs remain for treatment options with better efficacy and safety profiles, particularly among patients with refractory disease.

Activation of the type I IFN pathway plays a key role in SLE immunopathophysiology. Genome-wide association analysis has identified the genes encoding IFN regulatory factor 5, interleukin-1 receptor-associated kinase 1, and toll-like receptor 7 as disease susceptibility genes for SLE [43]. These are central molecules of the innate immune system, which is mainly composed of dendritic cells producing type I IFN [43].

Excessive suppression of the innate immune system, which plays a central role in the pathogenesis of SLE, could increase the risk of infection, as could targeting type I IFN or other cytokine pathways. Addressing adaptive, or acquired, immune system dysregulation associated with SLE through B-cell-targeted therapy was expected to become a mainstay of treatment for SLE, but biologics targeting B cells, such as rituximab and epratuzumab, have been unsuccessful [20]. Research has more recently focused on the inhibition of cytokines bridging the innate and acquired immune systems as a possible treatment approach for SLE. Successful treatment with biologics targeting cytokines produced by dendritic cells, such as soluble BAFF and type I IFN, which form a bridge between the innate and acquired immune systems, is of particular interest.

Results of clinical trials with anifrolumab have demonstrated the therapeutic benefit of inhibiting the IFN pathway in patients with SLE. Anifrolumab efficacy was observed as early as 8 weeks, when numerical separation of BICLA response rates by >10% was observed in favor of anifrolumab 300 mg and remained throughout 52 weeks of treatment. Early onset of clinical benefit (particularly in reducing severity of skin disease) is also demonstrated by the large proportions of anifrolumab-treated patients achieving ≥50% reduction in CLASI activity score from baseline to Week 12 compared with placebo in the TULIP-1, TULIP-2, and MUSE trials. Furthermore, anifrolumab demonstrated superiority to placebo in reducing OCS dosage.

The MUSE trial may have marked the first step toward precision medicine for patients with SLE. As SLE is a highly heterogeneous disease, it has been difficult to achieve similar therapeutic effects in all patients with the disease, for example, just by applying B-cell-targeted therapy to remove B cells [44]. However, a significant difference in therapeutic responsiveness to anifrolumab based on expression level of IFNGS was demonstrated in the MUSE and subsequent TULIP trials [35–37].

Future treatments will be based on subgroups of the disease at the cellular and molecular levels from information obtained from omics analyses, including immunophenotype [45]. Analysis of anifrolumab efficacy in different ethnicities is ongoing.

Conclusion

BICLA response was consistently achieved by a greater number of patients receiving anifrolumab compared with placebo across all three studies, as were sustained OCS reduction and improvement in CLASI. In both MUSE and TULIP-2, a greater percentage of patients also achieved an SRI(4) response with anifrolumab vs placebo. The safety profile of anifrolumab was generally similar across the efficacy studies, with SAEs occurring in 8–16% of anifrolumab-treated patients and 16–19% of patients receiving placebo. There was an increased incidence of herpes zoster in the anifrolumab treatment groups compared with placebo groups in all three studies, but most occurrences were cutaneous in presentation and responded to SOC treatment. Evidence from the clinical trials discussed suggests that in patients with active SLE, anifrolumab 300 mg administered IV Q4W is superior to placebo in achieving composite endpoints of disease activity response, as well as reducing OCS dosage, severity of skin disease, and flare rate.

Acknowledgments

The authors thank the patients, study sites, and investigators who participated in these clinical trials and also appreciate all members of the global academic steering committee and local steering committees. Writing and editing assistance was provided by Debra Scates, PhD, and Angela Cimmino, PharmD, of JK Associates Inc., a Fishawack Health Company. This support was funded by AstraZeneca.

Conflict of interest

Yoshiya Tanaka has received speaking fees and/or honoraria from Daiichi-Sankyo, Eli Lilly, Novartis, YL Biologics, Bristol Myers Squibb, Eisai, Chugai, AbbVie, Astellas, Pfizer, Sanofi, Asahi Kasei, GSK, Mitsubishi-Tanabe, Gilead, and Janssen, and has received research grants from Mitsubishi-Tanabe, Chugai, AbbVie, Takeda, UCB, Daiichi-Sankyo, and Eisai. Raj Tummala is an employee of AstraZeneca.