Safety and Tolerability of Omalizumab, A Randomized Clinical Trial of Humanized anti-IgE Monoclonal Antibody in Systemic Lupus Erythematosus (STOP LUPUS).

Abstract

BACKGROUND:

Autoreactive IgE antibodies have been implicated in the pathogenesis systemic lupus erythematosus (SLE). We hypothesized that omalizumab, a monoclonal antibody (mAb) binding IgE, may improve SLE activity by reducing type I IFN production by hampering plasmacytoid dendritic cells and basophil activation. This study assessed the safety, tolerability, and clinical efficacy of omalizumab in mild to moderate SLE.

METHODS:

Fifteen subjects with SLE and a Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI 2K) of > 4 and elevated autoreactive IgE antibodies were randomized to receive omalizumab or placebo (2:1) for 16 weeks, followed by 16-week open label treatment and 4-week washout period. SLEDAI 2K, British Isles Lupus Assessment Group index (BILAG 2004) and Physician Global Assessment (PGA) were recorded at each visit. Type I interferon (IFN) induced gene signature was determined using quantitative PCR.

RESULTS:

Omalizumab was well tolerated with no allergic reactions, and mostly mild adverse events comparable to placebo treatment. SLEDAI 2K scores improved in the omalizumab group at week 16 (p=0.038), as well as during the open label phase in subjects initially receiving placebo (p=0.020). No worsening in BILAG scores or PGA were detected. Omalizumab led to a trend towards reduction in IFN gene signature in subjects treated with omalizumab (p=0.11), especially in subjects with high baseline IFN signature (p=0.052).

CONCLUSION:

Omalizumab is well tolerated in SLE and associated with improvement in disease activity. Larger randomized clinical trials will be needed to assess efficacy of omalizumab in patients with SLE.

Introduction

The immune system in systemic lupus erythematosus (SLE) is characterized by abnormal activation of immune responses and the production of autoantibodies. While until recently most of the pathogenic autoantibodies in SLE were considered to be of the IgG subclass [1], we recently showed that Lyn and FcγRII deficient mice develop autoreactive IgE antibodies against anti-double stranded DNA (dsDNA) as well as lupus-like nephritis that is amplified by basophils in an IgE- dependent manner [2, 3]. SLE patients also have serum self-reactive IgEs and increased number of activated basophils in lymph nodes and spleens than can activate both T and B lymphocytes [2]. The presence of IgE antibodies against dsDNA, Sm and SSA was associated with active SLE and hypocomplementemia, suggesting that these autoantibodies may contribute to lupus pathogenesis [4, 5]. Furthermore, IgE autoantibodies do not correlate with total serum IgE levels and show only a modest correlation to serum IgG autoantibodies [5].

Additional studies have found that IgE immune complexes (IgE-IC) activate plasmacytoid dendritic cells (pDCs) upon binding cell surface Fc epsilon receptors (FcɛRI). These ICs induce type I IFN production in a TLR9-dependent manner [6]. When compared to IgG-IC, lower concentrations of IgE-IC are required to trigger proinflammatory responses. Hence, IgE- ICs may be involved in the maintenance and amplification of autoimmunity in SLE by inducing heightened activation of autoreactive immune responses [3, 6].

Omalizumab is a humanized IgG1 monoclonal antibody against human IgE that blocks the ability of IgE to bind FcɛRI. It is approved by the Food and Drug Administration (FDA) for the treatment of asthma and chronic idiopathic urticaria [7, 8] effectively depletes circulating IgE and decreases basophil activation [9, 10]. We hypothesized that depleting IgE autoantibodies with omalizumab in those SLE subjects with elevated autoreactive IgEs would hamper type I IFN production and reduce autoantibodies, possibly by reducing basophil and pDC activation, and could potentially reduce disease activity. We report the results of a phase Ib study designed to evaluate safety and tolerability of omalizumab in SLE. The study also assessed clinical efficacy and effects of IgE blockade on the type I IFN gene signature.

Methods

Trial design and interventions:

The study was approved by Institutional Review Board at the NIH and was conducted in accordance to the International Conference on Harmonization Good Clinical Practice requirements (Clinical Trials registration number: NCT01716312). The trial was monitored by an independent Data Safety Monitoring Committee for ongoing safety and pre-defined stopping rules. Participants were evaluated and all data collection occurred at the NIH Clinical Center in Bethesda, Maryland. The study was designed to be conducted in 3 phases (Figure S1). In the first double-blind phase, subjects were randomized at a 2:1 ratio to receive monthly subcutaneous omalizumab or placebo for 16 weeks. In the second open label phase, all subjects received subcutaneous omalizumab for 16 weeks. In the third phase, subjects were followed off study drug for 4 additional weeks. Subcutaneous omalizumab was administered as 600 mg loading dose followed by 300 mg every 4 weeks, based on simulation models created from data obtained from previous studies on the use of omalizumab in subjects with asthma and healthy volunteers. This dosing regimen can suppress more than 90% of circulating free IgE and significantly downregulates basophil high affinity FcɛRI along with membrane surface IgE [11]. The model was qualified as fit for purpose by confirming that simulations reproduced experimental data [12].

Participants:

Adult patients diagnosed with SLE fulfilling at least 4 of the criteria as defined by American College of Rheumatology [13] were eligible to participate. Additional inclusion criteria were: increased levels of (above 2 SD of the mean of healthy controls) IgE-anti-dsDNA, anti-Sm and/or anti-SSA autoantibodies measured by ELISA assay, as previously published [2]), and moderately active non-renal, non-CNS lupus, defined as Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI 2K) in the range of 4–14. Subjects were kept on stable background immunosuppressive regimen and prednisone ≤ 20 mg /day or equivalent throughout the study, with the option of one-time intramuscular injection of methylprednisolone for disease flare (defined as increase in SLEDAI 2K score by 5 or more compared to baseline). Any subject that continued to have disease activity (as determined by an increase in SLEDAI 2K by greater than 4 compared to study entry) despite one-time dose intramuscular methylprednisolone was withdrawn from the trial. Subjects were excluded if they were on biologics, had a history of asthma, anaphylaxis, known or suspected helminthic infection/infestation, weight >105 kg, total serum IgE level > 700 IU/mL, or serum creatinine > 2.0 mg/dL (Table S1).

Clinical Endpoints:

Adverse events (AE) were graded according to NCI Common Toxicity Criteria Version 4.0 (CTCAE 4.0) and relatedness of adverse events to intervention was determined according to pre-defined criteria. Disease activity measures included SLEDAI 2K and the British Isles Lupus Assessment Group (BILAG) 2004 index. Global disease activity was rated by the Physician Global Assessment (PGA). The severity of cutaneous involvement was documented on the Cutaneous Lupus Erythematosus Disease Area and Severity Index (CLASI). A 28 tender and swollen joint count as specified by Disease Activity Score in 28 joints (DAS28) was performed. Patient reported outcomes were measured using Visual Analog Scale (VAS), Multidimensional Assessment of Fatigue (MD-Fatigue) questionnaire and Short Form 36 (SF36).

Assessment of Type I IFN-gene signature (IFNGS):

The IFNGS was determined using a previously validated four gene score by assessing the levels of IFI27, IFI44, IFI44L, and RSAD2 through quantitative polymerase chain reaction (q-PCR) using TaqMan gene expression assay (ThermoFisher Scientific, Inc.) from subjects’ whole blood [14, 15]. Peripheral blood was collected by venipuncture in PAXgene Blood RNA tubes (BD Diagnostics, Inc) and stored at −20°C. RNA was isolated using PAXgene Blood RNA Kit (Qiagen, Inc.) following manufacturer’s instructions. Reverse Transcription Supermix for RT-qPCR by iScript (Bio-Rad, Inc) was used to synthesize cDNA. Expression level of GAPDH was used as housekeeping gene to normalize gene expression. ∆∆C_t was calculated by subtracting the mean determined from a set of healthy controls (n=30) for that particular gene. An IFN score was calculated as the average of the fold inductions for the 4 genes. High IFN score was determined as subjects with IFNGS above the two-SD of the mean of the healthy controls.

Statistical Methods:

This was a pilot, Phase Ib study to study the tolerability/toxicity of omalizumab, the number of subjects enrolled in this study was based primarily on the conventional number used in similar studies.

The baseline clinical characteristics were summarized using median with interquartile range (IQR) for continuous outcome and proportions for categorical outcomes and were compared using Wilcoxon rank sum test for continuous variables and exact Chi-square tests for categorical variables. Percent changes (or differences) from baseline of outcome of interest was calculated at each time point. Linear mixed model was conducted to take into account the correlation among repeated measures per subject, in order to assess difference between placebo and omalizumab using either percent changes or differences. Significant differences between treatment obtained using linear mixed model were summarized using estimated least squared means (LSmeans) with corresponding Standard errors (StdErr). For univariate analysis, comparison between placebo and omalizumab was conducted using Wilcoxon rank sum test. All analyses were two-tailed tests based on α = 0.05 and conducted using SAS Version 9.4 (SAS Institute, Cary, NC).

Results:

Subject characteristics

Sixty-two subjects were screened, 46 were excluded and 16 subjects were randomized (omalizumab n=10, placebo n=6; Figure 1, Table S2). Baseline demographics and disease activity were similar, except that subjects randomized to omalizumab had longer duration of disease compared to placebo subjects (p=0.02) and there were no Caucasians in the placebo group (Table S3). All subjects were on hydroxychloroquine and 80 % of subjects were on prednisone with the average prednisone dose of approximately 7 mg/day in both groups (Table S4).

Figure 1: CONSORT diagram.

Flow diagram of progress of patients through enrollment, intervention allocation and data analysis phases.

Safety Profile of Omalizumab in SLE

Omalizumab was well tolerated in patients with SLE. There were no localized or systemic allergic reactions in the double blind or open label phases. There were 52 AEs in 15 subjects during a span of 36 weeks. Most of these AEs (94.2%) were mild or moderate. In the first 16 weeks of treatment, subjects on omalizumab experienced 9 AEs, while those on placebo had 12 AEs (Table 1). No trend of involvement of a particular organ system was identified and there was no significant difference between treatment and placebo groups in terms of mild to moderate AEs.

Table 1: Adverse Events during double blind phase (day 0 to week 16) by treatment assignment, Body system, Preferred Term, and Greatest Severity by numbers and (percentages).

Adverse events (AE) were graded according to National Cancer Institute (NCI) Common Toxicity Criteria Version 4.0 (CTCAE 4.0).

Adverse Events
Body system	Omalizumab N=10	Placebo N=5	Total N=15
Blood and lymphatic system disorders
Mild	0(0.0%)	2(16.7%)	2(9.5%)
Moderate	0(0.0%)	2(16.7%)	2(9.5%)
Eye disorders
Mild	0(0.0%)	0(0.0%)	0(0.0%)
Moderate	0(0.0%)	1(8.3%)	1(4.7%)
Gastrointestinal disorders
Mild	0(0.0%)	0(0.0%)	0(0.0%)
Moderate	1(11.1%)	0(0.0%)	1(4.7%)
Infections and infestations
Mild	0(0.0%)	0(0.0%)	0(0.0%)
Moderate	2(22.2%)	0(0.0%)	2(9.5%)
Investigations
Mild	2(22.2%)	1(8.3%)	3(14.3%)
Moderate	0(0.0%)	0(0.0%)	0(0.0%)
Musculoskeletal disorders
Mild	0(0.0%)	2(16.7%)	2(9.5%)
Moderate	0(0.0%)	0(0.0%)	0(0.0%)
Nervous system disorders
Mild	1(11.1%)	0(0.0%)	1(4.7%)
Moderate	0(0.0%)	0(0.0%)	0(0.0%)
Renal and urinary disorders
Mild	0(0.0%)	0(0.0%)	0(0.0%)
Moderate	1(11.1%)	1(8.3%)	2(9.5%)
Respiratory disorders
Mild	1(11.1%)	2(16.7%)	3(14.3%)
Moderate	0(0.0%)	0(0.0%)	0(0.0%)
Skin disorders
Mild	1(11.1%)	0(0.0%)	1(4.7%)
Moderate	0(0.0%)	0(0.0%)	0(0.0%)
Vascular disorders
Mild	0(0.0%)	0(0.0%)	0(0.0%)
Moderate	0(0.0%)	1(8.3%)	1(4.7%)
Total
Mild	5(55.6%)	7(58.3%)	12(57.1%)
Moderate	4(44.4%)	5(41.7%)	9(42.9%)

Three AEs categorized as serious occurred during the study. One subject, while on placebo, developed chest pain secondary to bronchitis and cough and required overnight admission for observation to rule out acute coronary syndrome. Another subject randomized to omalizumab developed primary varicella infection at week 18 following exposure to chicken pox. This individual had emigrated from West Africa to the United States, had no evidence of prior immunity against varicella and was on azathioprine 2mg/kg/day. One subject initially randomized to placebo developed pulmonary embolism after starting the open phase of the study at week 16. Evaluation was negative for antiphospholipid antibodies and other genetic or acquired risk factors for thromboembolic events, raising the possibility of this event being related to omalizumab. One patient was withdrawn from study by the investigators after randomization, as her underlying interstitial lung disease was considered an unacceptable risk for a phase I safety study. Another patient with history of lupus nephritis was withdrawn at week 28 as she met the withdrawal criteria of > 50 % worsening of proteinuria (protein to creatinine ratio increased from baseline level of 0.6 mg/mg to 1.5 mg/mg).

Changes in SLE disease activity during the study

At week 16, there was improvement in SLEDAI 2K scores in the omalizumab group compared to the placebo group (p=0.038). This improvement was maintained until the end of the treatment period (week 32) with a trend toward worsening SLEDAI 2K scores (p=0.54) after 4 weeks off drug (Figure 2 and S2). Disease severity in the group of 5 subjects assigned to placebo during the first 16 weeks of the trial also showed improvement in the SLEDAI 2K after subsequently being placed on omalizumab for 16 weeks during the open label phase (p=0.02). While changes in SLEDAI 2K were statistically significant, the average reduction of score by 2 points may not be clinically meaningful (Table S5). The improvements in SLEDAI 2K were mainly in arthritis, rash and serological domains. Three subjects achieved composite response criteria of SRI 4: two subjects randomized to omalizumab achieved SRI 4 at week 16 and maintained it at Week 32, while one subject on placebo during the first 16 weeks met the SRI 4 after open label treatment at week 32. There were no significant changes observed in patient reported outcomes SF-36 and MD-Fatigue. Disease activity as marked by subjects on a visual analog scale (VAS) remained stable in omalizumab group (Figure S3). Steroid taper or withdrawal of immunosuppressive medications was not attempted during the study.

Figure 2: Median SLEDAI 2K scores during the double-blind phase (weeks 0–16).

Vertical bars represent the Interquartile ranges (IQRs) of the observed data. SLEDAI 2K: Systemic Lupus Erythematosus Disease Activity Index 2000. Week 16 * p = 0.038

SLE disease activity lab values (IgG anti-dsDNA, C3 and C4) remained stable throughout the intervention and wash out phase (Figures S4, S5 and S6). As expected due to measurement of circulating drug-IgE complexes, median total serum IgE levels increased after starting omalizumab in subjects initially randomized to omalizumab and in subjects initially randomized to placebo after they were started on omalizumab during the open label phase (18).

Effects of omalizumab on IFNGS

Subjects receiving omalizumab showed a trend toward improvement in IFNGS during the first 16 weeks of therapy (Figure S7). This difference was more pronounced in subjects who had high IFNGS at baseline (Figure S8). This suggests that omalizumab may modulate type I IFN pathways by blocking self-reactive IgE, and is consistent with previous reports of stimulation of type I IFN in pDCs by IgE ICs [6].

Discussion

The identification of IgE autoantibodies and their role in basophil activation in animal models and in human SLE has revealed a new pathway potentially involved in SLE pathogenesis [2, 5]. In this proof-of-concept study, omalizumab, a humanized IgG1 monoclonal antibody against human IgE, was tested for the first time as an add-on therapy in SLE. A potential advantage of omalizumab in SLE is a side effect profile different from immunosuppressive drugs and a convenient once a month subcutaneous administration schedule. We have previously reported presence of IgE autoantibodies in greater than 70% of SLE patients, hence omalizumab as an add-on therapy to standard of care may be applicable to a large subset of SLE patients [5].

In this pilot clinical trial omalizumab was well tolerated with no immediate or delayed allergic reactions. Statistically significant improvement in disease activity was detected after subjects received omalizumab for 16 weeks and 32 weeks by measuring change from baseline in SLEDAI 2K score, but not in other measures of clinical activity. The proportion of SRI-4 responders was low when compared to what has been reported in other larger clinical trials. This is probably due to the small sample size studied and the overall mild disease activity at enrollment. In addition, the absolute change in the SLEDAI 2K score was approximately 2, which is below what is usually considered a clinically minimally important change [16, 17]. There was no worsening in patient reported outcomes while subjects were on omalizumab.

A five-year observational study of omalizumab in subjects with asthma has suggested a potential increased risk of serious cardiovascular and cerebrovascular events [18]. In the current study, one subject developed pulmonary embolism after receiving the first dose of omalizumab. This raises the concern of whether the event was related to omalizumab or due to increased background risk of thromboembolic events in patients with SLE. Given that CV risk is enhanced in SLE [19], future studies using omalizumab in SLE should include markers of vascular risk and assessments of vascular dysfunction.

This study is the first to show that omalizumab reduced IFNGS, especially in SLE subjects with high IFNGS at baseline. This suggests that the mechanism of action of this drug in SLE may be at least in part through abrogation of dysregulated IFN pathways. In future studies, other biomarkers potentially involved in this pathway such as CD62L and Th2 cytokines should be measured.

Our study has several limitations. Importantly, we studied a small number of patients with mild clinical disease with no organ threatening manifestations. The safety data from this study needs to be confirmed in a larger cohort with more active disease, by excluding subjects with elevated risk of thromboembolic events. This exploratory study was not powered for efficacy and a larger sample size will be needed to ascertain if omalizumab has a potential to treat a selected subset of patients with SLE and high levels of anti-IgE antibodies. The mechanism of action of omalizumab in SLE also needs to be better defined by performing additional mechanistic studies controlling for inter-test variability.

Overall, this is the first trial to test the safety and potential efficacy of blocking IgE autoantibodies as a novel non-immunosuppressive agent in treatment of SLE. The results from this exploratory study need to be confirmed in future larger studies adequately powered to measure efficacy.