Down-regulation of interferon signature in systemic lupus erythematosus patients by active immunization with interferon α–kinoid

Authors


  • ClinicalTrials.gov identifier: NCT01058343.

Abstract

Objective

We developed interferon-α–kinoid (IFN-K), a drug composed of inactivated IFNα coupled to a carrier protein, keyhole limpet hemocyanin. In human IFNα–transgenic mice, IFN-K induces polyclonal antibodies that neutralize all 13 subtypes of human IFNα. We also previously demonstrated that IFN-K slows disease progression in a mouse model of systemic lupus erythematosus (SLE). This study was undertaken to examine the safety, immunogenicity, and biologic effects of active immunization with IFN-K in patients with SLE.

Methods

We performed a randomized, double-blind, placebo-controlled, phase I/II dose-escalation study comparing 3 or 4 doses of 30 μg, 60 μg, 120 μg, or 240 μg of IFN-K or placebo in 28 women with mild to moderate SLE.

Results

IFN-K was well tolerated. Two SLE flares were reported as serious adverse events, one in the placebo group and the other in a patient who concomitantly stopped corticosteroids 2 days after the first IFN-K dose, due to mild fever not related to infection. Transcriptome analysis was used to separate patients at baseline into IFN signature–positive and –negative groups, based on the spontaneous expression of IFN-induced genes. IFN-K induced anti-IFNα antibodies in all immunized patients. Notably, significantly higher anti-IFNα titers were found in signature-positive patients than in signature-negative patients. In IFN signature–positive patients, IFN-K significantly reduced the expression of IFN-induced genes. The decrease in IFN score correlated with the anti-IFNα antibody titer. Serum complement C3 levels were significantly increased in patients with high anti-IFNα antibody titers.

Conclusion

These results show that IFN-K is well tolerated, immunogenic, and significantly improves disease biomarkers in SLE patients, indicating that further studies of its clinical efficacy are warranted.

Systemic lupus erythematosus (SLE) is a severe chronic autoimmune disease that primarily affects young women and causes inflammation and injury to multiple organs (1). SLE is characterized by B cell hyperreactivity and the loss of tolerance to chromatin, resulting in the production of autoantibodies against nuclear targets, such as double-stranded DNA (dsDNA) (2). SLE is traditionally treated with antimalarials, corticosteroids, and other immunosuppressive agents. These drugs reduce mortality, but not all patients respond equally to these treatments, and their use can lead to serious side effects, such as osteoporosis, bone marrow suppression, and infection.

Interferon-α (IFNα) plays a critical role in the pathogenesis of SLE, in particular by increasing the (auto)antigen-presenting abilities of monocytes/dendritic cells, thereby leading to the activation of autoreactive T cells (3). Several studies have demonstrated an overexpression of IFNα-inducible genes in SLE patients and a correlation between their expression and disease severity (4–7). Accordingly, several IFNα-blocking monoclonal antibodies are being developed for the treatment of SLE. A recent phase I study showed that sifalimumab, a fully human IFNα-blocking monoclonal antibody, is well tolerated and may improve clinical scores in SLE patients (8). However, unlike polyclonal antibodies, monoclonal antibodies are unlikely to neutralize all 13 IFNα subtypes. Also, monoclonal antibodies are expensive and difficult to produce.

We are currently developing IFNα–kinoid (IFN-K) as an alternative treatment for SLE. Kinoids are composed of inactivated cytokines conjugated to a carrier protein, keyhole limpet hemocyanin (KLH), and are then injected as an emulsion with an adjuvant. In animal studies, immunization with kinoids induces high titers of polyclonal neutralizing antibodies against the targeted cytokine (9). IFN-K slows disease progression in a mouse model of SLE (10), and in human IFNα–transgenic mice, it induces polyclonal antibodies that neutralize all 13 subtypes of human IFNα but not IFNβ or IFNγ (11). These findings suggest that immunization with IFN-K will induce the production of antibodies to IFNα in SLE patients and thereby reduce disease activity.

Herein we describe the results of a phase I/II dose-escalation study designed to assess the safety, immunogenicity, and biologic effects of IFN-K in adults ages 18–50 years with SLE. We also examined the effect of IFN-K on the expression of IFNα-inducible genes, which have previously been described as a robust marker of SLE disease activity (6, 12).

PATIENTS AND METHODS

Study design.

This was a multicenter, randomized, double-blind, placebo-controlled, phase I/II staggered dose-escalation study of IFN-K in adults with SLE (ClinicalTrials.gov registry number NCT01058343). The study was performed between May 6, 2010 and August 3, 2011. All aspects of the study were carried out in accordance with the Declaration of Helsinki (as revised by the 59th World Medical Association General Assembly, Seoul, South Korea, October 2008), the European Medicines Agency Guidelines on Clinical Evaluation of New Vaccines (13), and the International Conference for Harmonization, Topic E6: Good Clinical Practice Guidelines (14). Institutional ethics committee approval was obtained for all centers. All subjects provided written informed consent before participating in this study.

SLE patients.

Adults ages 18–50 years with a diagnosis of mild to moderate SLE were recruited for this study. Mild to moderate SLE was defined as the presence of at least 4 of the 11 American College of Rheumatology criteria for SLE (15), an SLE Disease Activity Index 2000 (SLEDAI-2K) score (16) between 4 and 10, and the presence of serum antinuclear antibodies, anti-dsDNA antibodies, or both. Patients with severe disease manifestations (cytopenia, renal involvement, or central nervous system involvement) were excluded in order to avoid the use of strong immunosuppressive therapies that would decrease the ability to detect significant differences between IFN-K–treated patients and placebo-treated patients. Further inclusion and exclusion criteria are shown in Supplementary Table 1A, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37785/abstract.

Treatments.

SLE patients were treated with IFN-K or an equivalent volume of 0.9% NaCl (placebo), both mixed with an equal volume of Montanide ISA-51vg adjuvant (Seppic). According to the manufacturer, Montanide ISA-51vg is composed of light mineral oil and a surfactant system designed to make a water-in-oil emulsion. IFN-K was prepared by crosslinking recombinant human IFNα2b with KLH as described previously (11). Patients were randomized to treatments using an interactive web response system, and both patients and investigators were blinded to the treatment type. All treatments were delivered by intramuscular injection. Patients were randomized to receive placebo, 30 μg IFN-K, 60 μg IFN-K, 120 μg IFN-K, or 240 μg IFN-K, as shown in Figure 1.

Figure 1.

Study flow and patient disposition. In group 1, 4 patients with systemic lupus erythematosus were randomized 3:1 to receive 30 μg interferon-α–kinoid (IFN-K) or placebo. After safety data were reviewed, the 8 patients in group 2 were randomized 3:1 to receive 60 μg IFN-K or placebo. The same safety review procedure was followed prior to escalation to the next-higher dose group, in which 6 patients received 120 μg IFN-K and 2 patients received placebo, and again for the final group, in which 6 patients received 240 μg IFN-K and 2 patients received placebo. All groups received an additional dose of IFN-K or placebo on day 7 and on day 28. On day 84, patients in the 60 μg IFN-K, 120 μg IFN-K, and 240 μg IFN-K groups received a fourth dose of IFN-K or placebo if investigators decided there were no safety issues from the third dose of IFN-K, and if day-56 serum samples from at least 2 of the first 4 patients in the group did not contain human IFNα–neutralizing antibodies. AE = adverse event.

Clinical assessments.

The SLEDAI-2K (16) and the British Isles Lupus Assessment Group (BILAG) 2004 scores (17) were determined by investigators at enrollment (day 0) and on days 56, 112, and 168.

Safety.

Adverse events (AEs) were defined and their occurrence, intensity, and relationship to IFN-K immunization were recorded during the study, as described in the European Medicines Agency Guidelines on Clinical Evaluation of New Vaccines (13) and International Conference for Harmonization, Topic E6: Good Clinical Practice Guidelines (14). Solicited local reactions (pain, tenderness, erythema, swelling, itching, induration, and ulceration) and systemic reactions (fever, vomiting, headache, fatigue, myalgia, and nausea) were recorded by patients on diary cards for 7 days following each injection of IFN-K or placebo. Unsolicited AEs and serious AEs (SAEs), as well as all abnormalities found on physical examination, vital signs, 12-lead electrocardiogram results, and findings of clinical laboratory evaluations, were recorded by investigators throughout the study.

Measurement of anti-KLH and anti-IFNα antibodies.

Anti-IFNα and anti-KLH antibody titers were measured in sera by enzyme-linked immunosorbent assay as described previously (11). Briefly, plates were coated with KLH or IFNα and incubated with serum, and bound antibodies were detected using horseradish peroxidase–conjugated goat anti-human immunoglobulin (AbD Serotec). The anti-IFN titers for each sample were expressed as 1:n, where n is the highest dilution for which the mean optical density (at 492 nm) was higher than the cutoff value based on a pool of sera from healthy volunteers. Anti-KLH titers were expressed as 1:n, where n is the highest dilution for which the mean optical density was twice the optical density of the preimmune sample at a dilution of 1:200. The lower limit of quantitation for both assays was 200, which was the lowest dilution of serum tested.

Measurement of IFNα-neutralizing capacity.

IFNα-neutralizing capacity was measured as previously described, using a bioassay that assesses the antiviral activity of human IFNα in Madin-Darby bovine kidney cells in the presence of vesicular stomatitis virus (11). The 50% capacity for neutralization of IFN activity (10 units/ml) was the dilution giving 50% of the maximum signal according to least squares linear regression for the linear portion of the curve. The lower limit of quantitation was 200, which was the lowest dilution of serum tested (1:200).

Cellular response assay.

Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque density-gradient centrifugation, added (200,000/well) to round-bottomed 96-well plates, and incubated at 37°C with IFN-K, recombinant human IFNα2b (European Directorate for the Quality of Medicines), KLH (Stellar Biotechnologies), glutaraldehyde–formaldehyde–treated IFNα, glutaraldehyde–formaldehyde–treated KLH at the indicated concentrations, or medium alone. Glutaraldehyde–formaldehyde–treated IFNα was generated as described previously (11). Glutaraldehyde–formaldehyde–treated KLH was generated in the same way as glutaraldehyde–formaldehyde–treated IFNα but using KLH in place of IFNα. After 72 hours, 3H-thymidine incorporation was assessed as described previously (11). All samples were tested in quadruplicate. The proliferation index was calculated as the counts per minute for stimulated cells divided by the counts per minute for cells cultured in medium alone.

Measurement of serum complement C3 and C4 and serum anti-dsDNA antibody concentrations.

Serum complement C3 and C4 levels were measured by immune complex formation assays (Beckman Coulter). For C3, the detection range was 5.83–12,600 mg/dl, and the reference range for healthy individuals was 79–152 mg/dl. For C4, the detection range was 1.67–4,680 mg/dl, and the reference range for healthy individuals was 16–38 mg/dl. Anti-dsDNA antibody titers were measured by radioimmunoassay (Diagnostic Products). The detection range for dsDNA was 2.5–50 IU/ml, and the reference range for healthy individuals was 0–5.3 IU/ml.

Gene expression profiling.

Whole blood samples obtained from all SLE patients (on days 0, 38, 112, and 168) and from 50 healthy volunteers were collected in PAXgene Blood RNA tubes (Qiagen). Healthy volunteers had to be 18–50 years of age. Lactating or pregnant women were excluded, as were patients who had an elevated serum C-reactive protein level or were positive for antinuclear antibodies, anti-dsDNA antibodies, or rheumatoid factor. An additional whole blood sample was collected from 18 of the healthy volunteers. Aliquots of these 18 whole blood samples were added to Na-heparin tubes prefilled with phosphate buffered saline. In addition, aliquots were added to Na-heparin tubes prefilled with a mixture of all 13 IFNα subtypes (PBL InterferonSource; 3 × 50% maximum response concentration [EC50]/ml each) for 9 of the samples, or with IFNα2b only (3 × EC50/ml) for the other 9 samples. The samples were incubated for 4 hours at 37°C and then transferred to PAXgene Blood RNA tubes.

RNA was extracted from blood samples (control samples, IFN-stimulated control samples, and samples obtained from SLE patients on days 0, 112, and 168) using a PAXgene Blood RNA kit according to the manufacturer's instructions (Qiagen). RNA quality was assessed using an Agilent 2100 Bioanalyzer and RNA nanochips. Samples with an RNA integrity number of <6 were discarded. RNA was labeled using a One-Cycle Target Labeling kit (Affymetrix) and hybridized to GeneChip human genome U133 Plus 2.0 arrays (Affymetrix). The slides were washed and stained on a GeneChip Fluidics Station (Affymetrix) and scanned using a GeneChip Scanner 3000 (Affymetrix). The Affymetrix CEL files were deposited in the Gene Expression Omnibus (GEO) of the National Center for Biotechnology Information, and are accessible through GEO accession number GSE39088.

One SLE patient was excluded from gene expression analysis because of a missing baseline sample. All of the analyses described in this report were repeated using extrapolated baseline data for this patient (based on expression data of a day-38 whole blood sample), and the same results were obtained. Also, an SLE patient who had a disease flare before receiving a second dose of IFN-K was excluded from the analysis on days 112 and 168. Baseline analyses, performed after exclusion of this patient, delivered the same results as those described in this report. One placebo-treated patient received high-dose oral and intravenous steroids between days 112 and 168, so her day-168 samples were excluded from the analysis. Thus, valid gene expression profiles were obtained for 27 SLE patients at baseline, 26 on day 112, and 25 on day 168. Valid gene expression profiles were also obtained for 46 of the 50 healthy volunteers (4 with a RNA integrity number of <6 were excluded) and for all 18 of the healthy samples that were stimulated with IFNα.

Statistical analysis.

No formal sample size calculations were performed for this study. Calculations were made for all patients who received at least 1 treatment. Prism 5 and Microsoft Excel were used. For calculations of the geometric mean titers (GMT) for the immunogenicity and neutralizing capacity analysis, titers <200 were considered to be 100, which was one-half the lower limit of quantitation. For gene expression profiling, fluorescence intensity data were normalized by robust multiarray analysis (18) using GeneSpring (Agilent). Unsupervised hierarchical clustering algorithms and t-tests were performed using GeneSpring. The IFN score for each sample was calculated using Microsoft Excel as the fold change between the median of the normalized expression of the 31 probe sets originated from the 21 IFN-inducible genes described by Yao et al (6) in the sample, and the average of the median values in the healthy volunteers. P values less than 0.05 were considered significant.

RESULTS

Characteristics of the SLE patients.

The study included 28 women with mild to moderate SLE with a mean ± SD age of 37.1 ± 9.3 years (range 19–50 years) and a range of SLEDAI-2K scores of 4.0–10.0. All patients were Caucasian, except for one Asian patient in the placebo group. Mean age, SLEDAI-2K scores, concomitant therapy, and disease duration were similar in all groups (see Supplementary Table 1, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37785/abstract. Twenty-one of the patients were treated with IFN-K, and 7 were treated with placebo (Figure 1). All 28 patients completed the study, although 1 patient in the 240 μg IFN-K group received only a single dose of IFN-K because of a disease flare.

IFN-K immunogenicity.

At baseline, 1 IFN-K–treated patient had low levels of anti-IFNα antibodies. None of the other patients included in the study had detectable anti-IFNα antibodies, anti-KLH antibodies, or IFNα-neutralizing activity at baseline. In 2 of the 7 placebo-treated patients, low levels of autoantibodies to IFNα were transiently detected at later time points.

In contrast, anti-IFNα antibodies were detected in patients from each group treated with IFN-K starting on day 17 (Figure 2A). GMTs tended to increase over time in all IFN-K treatment groups, and reached a peak between day 56 and day 168. GMTs were generally dose dependent, and were also higher in patients immunized with 4 doses than in patients immunized with 3 doses of IFN-K, although the difference was not statistically significant. The kinetics of antibody production and the effects of boosting were similar for anti-KLH antibodies (Figure 2B). In addition, IFNα-neutralizing activity was detected in 3 of 6 patients treated with 60 μg IFN-K, 3 of 6 patients treated with 120 μg IFN-K, and 4 of 5 patients treated with 240 μg IFN-K, but not in patients treated with 30 μg IFN-K (Figure 2C) or patients treated with placebo.

Figure 2.

Anti–human interferon-α (anti–human IFNα) antibody titers, anti–keyhole limpet hemocyanin (anti-KLH) antibody titers, and IFNα-neutralizing capacity in samples obtained from systemic lupus erythematosus patients treated with placebo or the indicated doses of IFNα-kinoid (IFN-K). A and B, Serum anti–human IFNα (A) and anti-KLH (B) antibody titers, determined by enzyme-linked immunosorbent assay on days 0, 17, 28, 38, 56, 112, 168, 252, and 336. Values are 1:geometric mean titer (1:GMT) in 1–3 patients per treatment group. Values <1:200 were considered to be 1:100. C, IFNα-neutralizing capacity of sera tested in vitro using infected cultured cells exposed to IFNα. NC50 = concentration giving half-maximal inhibition.

We also evaluated the magnitude of the cellular responses against IFN-K and its components on PBMCs obtained preimmunization and on days 56 and 112. Lymphoproliferation assays indicated that PBMCs from IFN-K–treated patients displayed significant proliferative responses when exposed to IFN-K itself, KLH, or glutaraldehyde–formaldehyde–treated KLH but not when exposed to IFNα or glutaraldehyde–formaldehyde– treated IFNα, as compared to placebo-treated patients (see Supplementary Figure 1, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37785/abstract.

Baseline IFN-inducible gene expression and correlation with disease severity and IFN-K immunogenicity.

High-density transcriptome studies, which were performed on whole blood samples obtained from all patients at baseline, confirmed that many genes that are up-regulated in SLE are IFN inducible. However, further analyses revealed that baseline gene expression profiles were not homogenous among SLE patients. Thus, unsupervised clustering analyses, performed on all baseline samples using the genes differentially expressed between SLE patients and controls, separated the patients into 2 groups. In the first group, expression of the 21 IFN signature genes was up-regulated compared to healthy volunteers (IFN signature–positive patients; n = 18), whereas in the second group, expression of these genes was similar to that in healthy volunteers (IFN signature–negative patients; n = 9) (Figure 3A).

Figure 3.

Baseline interferon (IFN) gene expression signature in systemic lupus erythematosus (SLE) patients and its effect on the induction of anti-IFNα and anti–keyhole limpet hemocyanin (anti-KLH) antibodies in response to IFNα-kinoid (IFN-K). A, Unsupervised hierarchical clustering. Eighteen SLE samples and 1 control sample clustered in a branch neighboring the IFN-stimulated control samples, which was distinct from a branch containing the remaining 9 SLE samples. The robust multiarray analysis–normalized expression of the 31 probe sets encoding the 21 IFN-induced genes that define the IFN score is shown for each sample. B and C, Anti-IFNα (B) and anti-KLH (C) antibody titers on day 112 and day 168 after the first injection in patients treated with placebo (n = 7), IFN signature–negative patients treated with IFN-K (n = 7), and IFN signature–positive patients treated with IFN-K (n = 12). Symbols represent individual patients; horizontal lines represent the mean. P values were determined by Mann-Whitney test.

As expected, patients with a positive IFN signature at baseline had higher biologic indices of disease severity as compared to IFN signature–negative patients. Thus, SLE patients with a positive IFN signature had significantly higher dsDNA antibody concentrations, lower complement C3 concentrations, and lower complement C4 concentrations than IFN signature–negative patients, although baseline SLEDAI-2K scores were not significantly different between the 2 groups (see Supplementary Table 2, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37785/abstract.

Given the stimulatory effects of type I IFN on antigen-presenting cells, we investigated whether production of anti-KLH and anti-IFNα antibodies in response to IFN-K differed between patients who had the IFN signature at baseline and those who did not. Notably, IFN-K induced significantly higher anti-IFNα (Figure 3B) and anti-KLH (Figure 3C) antibody titers in IFN signature–positive patients than in IFN signature–negative patients.

Effect of active immunization with IFN-K on IFN-inducible gene expression.

We compared the changes induced by IFN-K versus placebo in the expression of the 21 IFN signature genes in patients who had a positive IFN signature at baseline and those who did not. In IFN signature–positive patients, the decrease in IFN-inducible gene expression from baseline was significantly greater after IFN-K treatment than after placebo treatment, on both day 112 and day 168. As expected, IFN-K did not decrease IFN-inducible gene expression more than placebo in IFN signature–negative patients (Figure 4A).

Figure 4.

Effect of interferon-α–kinoid (IFN-K) on the expression of IFN-induced genes. A, Change from day 0 in the expression of the 31 probe sets encoding the 21 IFN signature genes in systemic lupus erythematosus patients treated with IFN-K or placebo. The differences in the average log2 expression of each of these probe sets between the indicated day and day 0 were calculated for patients treated with placebo (n = 7), IFN signature–negative patients treated with IFN-K (n = 7), and IFN signature–positive patients treated with IFN-K (n = 12). The distribution of these differences is shown for each group. Data are shown as box plots, where the boxes represent the interquartile range, the lines inside the boxes represent the median, and whiskers represent the minimum and maximum. ∗∗∗ = P < 0.001 by analysis of variance and Tukey's post hoc test. B, Correlation between anti-IFNα antibody titer on day 112 and the change in IFN score between day 112 and day 0 for all patients (Spearman's correlation r = −0.48, P = 0.013).

As shown in Supplementary Figure 2 (available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37785/abstract, the changes observed in IFN scores varied from patient to patient. We did not observe a significant effect of IFN-K dose or number of injections, probably because of the small numbers of patients included in each group. In contrast, we found a significant correlation between anti-IFNα antibody titer on day 112 and the effect on the IFN score. The higher the anti-IFNα antibody titer, the greater the decrease in score was between baseline and day 112 (r = −0.48, P = 0.013) (Figure 4B). Results were similar on day 168, although the trend was not statistically significant (r = −0.37, P = 0.079) (data not shown).

Since production of anti-IFNα antibodies was higher in patients with a positive IFN signature at baseline, we examined whether there was a link between IFN score at baseline and decrease in IFN score upon IFN-K therapy. Accordingly, we found that higher IFN scores at baseline were significantly associated with a greater decrease in IFN scores on day 112 in IFN-K–treated patients (r = −0.46, P = 0.046), but not in placebo-treated patients (r = 0.11, P = 0.84) (data not shown).

Clinical efficacy and safety.

Compared to day 0, SLEDAI-2K scores decreased in all groups, including the placebo group, so that no differences were detected between the different groups. Results were similar for the BILAG indices (data not shown). Changes from baseline in serum complement C3, complement C4, or anti-dsDNA concentrations were not significant between groups (data not shown). However, we found that patients who had high anti-IFNα titers (>1:25,600) on day 112 had significantly larger increases in serum C3 concentrations than patients who had lower titers (<1:25,600; P = 0.027) (Figure 5A). Serum C4 concentrations followed a similar although nonsignificant trend (Figure 5B). This trend was not observed for anti-dsDNA antibodies (Figure 5C).

Figure 5.

Effect of interferon-α–kinoid (IFN-K)–induced anti-IFNα antibodies on serum complement C3 levels (A), complement C4 levels (B), and double-stranded DNA (dsDNA) antibody titers (C) on day 112 in patients with systemic lupus erythematosus. The absolute changes in serum complement C3 levels, serum complement C4 levels, and serum dsDNA titers from day 0 to day 112 are shown for patients with low (<1:25,600) and high (>1:25,600) anti-IFNα antibody titers. Serum complement C3 and C4 levels were measured by immune complex formation assays, anti-dsDNA antibody titers were measured by radioimmunoassay, and anti-IFNα antibody titers were measured by enzyme-linked immunosorbent assay. Symbols represent individual patients; horizontal lines represent the mean. P values were determined by Mann-Whitney test.

All solicited injection site or systemic reactions were of mild or moderate severity (Table 1) (see Supplementary Table 3, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37785/abstract, for further details regarding solicited reactions). Only 2 SAEs were reported during the study. Both were SLE flares. One occurred 2 months after the last injection in a patient who was receiving placebo. The second was in a patient who had a flare 10 days after receiving the first dose of 240 μg IFN-K. This patient was withdrawn from the study by the investigator. She later reported that she had mistakenly stopped taking corticosteroids 2 days after IFN-K administration due to fever (38°C) and concerns of infection without informing the investigator. However, comprehensive laboratory analysis did not detect any bacterial or viral infections.

Table 1. Adverse events, solicited injection site reactions, and solicited systemic reactions between day 0 and day 6 after any treatment dose, and urinalysis results, for SLE patients in the phase I/II study of IFN-K*
 30 μg IFN-K (n = 3)60 μg IFN-K (n = 6)120 μg IFN-K (n = 6)240 μg IFN-K (n = 6)Any IFN-K dose (n = 21)Placebo (n = 7)
  • *

    Except where indicated otherwise, values are the number (%) of patients.

  • One patient in the 240 μg dose group withdrew from the study after the first injection of interferon-α–kinoid (IFN-K) due to a systemic lupus erythematosus (SLE) flare and was not included in the subsequent analyses.

  • Infections in the IFN-K–treated patients were upper respiratory tract infection (n = 2), bronchitis (n = 1), urinary tract infection (n = 3), sinusitis (n = 1), gastroenteritis (n = 1), herpes zoster (n = 1), and influenza (n = 1). Infections in the placebo-treated patients were nasopharyngitis (n = 1) and vulvovaginal mycotic infection (n = 1).

  • §

    Defined as >5 red blood cells/high-power field.

Serious adverse events (SLE flare)0 (0.0)0 (0.0)0 (0.0)1 (16.7)1 (4.8)1 (14.3)
Other adverse events (infection)0 (0.0)2 (33.3)3 (50.0)5 (83.3)10 (47.6)2 (28.6)
Any solicited injection site reaction2 (66.7)5 (83.3)5 (83.3)5 (83.3)17 (81.0)4 (57.1)
Any solicited systemic reaction2 (66.7)3 (50.0)4 (66.7)5 (83.3)14 (66.7)2 (28.6)
Urinalysis results      
 Protein/creatinine ratio, mean ± SD      
  Day 05.45 ± 5.643.14 ± 5.321.00 ± 0.9531.49 ± 74.9110.96 ± 39.912.07 ± 3.60
  Day 566.07 ± 7.791.58 ± 1.660.71 ± 0.2629.73 ± 69.4110.01 ± 37.1114.81 ± 34.74
  Day 1126.72 ± 10.231.41 ± 1.660.80 ± 0.285.64 ± 8.203.08 ± 5.6821.78 ± 55.88
  Day 1689.08 ± 14.742.20 ± 2.600.83 ± 0.6017.92 ± 41.837.29 ± 22.7131.47 ± 81.63
 Hematuria, no. of patients§      
  Day 0000000
  Day 56010010
  Day 112101131
  Day 168010010

DISCUSSION

This phase I/II study showed that IFN-K is well tolerated, induces anti-IFNα antibodies, and down-regulates IFN-induced genes in SLE patients overexpressing IFN-inducible genes at baseline. SLE patients with a positive IFN signature produced significantly higher titers of anti-IFNα antibodies in response to IFN-K compared to IFN signature–negative patients. The IFN-K–induced down-regulation of the IFN score correlated with the titers of anti-IFNα antibodies. Finally, this study also showed that the anti-IFNα antibody response induced by IFN-K is associated with the recovery of serum complement C3 levels.

Kinoid administration was well tolerated. The single SAE reported for an SLE patient treated with IFN-K was a disease flare, likely linked to abrupt stopping of corticosteroids. The only dose/volume-dependent AEs were injection site erythema and induration, and these were all mild to moderate in severity. SLE patients are prone to viral infections (e.g., influenza, cytomegalovirus, Epstein-Barr virus, and herpes). However, no severe viral infection was documented during the trial. This is potentially related to the redundancy of the IFN system, which uses IFNα, but also IFNβ or type III IFNs to fight viral infections. On the whole, these data suggest that IFN-K treatment is not associated with major toxicity.

IFN-K induced antibodies against IFNα in all SLE patients, sometimes at very high titers. Our previous studies of human IFNα–transgenic and NZB/W F1 lupus-prone mice showed that IFNα induces anti-IFNα antibodies only when it is conjugated to KLH (10, 11). In the present study, we observed a strong lymphoproliferative response to KLH in immunized patients, while this was not the case when PBMCs were exposed to glutaraldehyde–formaldehyde–treated IFNα or native IFNα. Taken together, our results suggest that the IFN-K–induced rupture of B cell tolerance to IFNα and production of anti-IFNα antibodies mainly rely on the activation of anti-KLH Th cells that provide bystander help to anti-IFNα B cells.

In this respect, the observation that patients overexpressing IFN-induced genes at baseline produced significantly higher (up to 10 times higher) titers of anti- KLH and anti-IFNα antibodies in response to IFN-K compared to IFN signature–negative patients is of particular interest. It provides further evidence of the purported role of type I IFNs in SLE, resulting in the priming of antigen-presenting cells and leading to improved responses to immunization. IFN signature–positive patients comprised approximately two-thirds of all SLE patients included in this study, which is consistent with the results of Yao et al (6), who reported a moderate or high IFN signature in 80% of SLE patients. Compared to IFN signature–negative patients, the IFN signature–positive patients included in our study had more active SLE disease at baseline, as indicated by lower complement C3, lower complement C4, and higher anti-dsDNA titers, consistent with the findings of several other studies that have examined the association between IFNα-inducible gene expression and SLE disease activity (4, 5, 7).

Morimoto et al (19) recently reported that 27% of serum samples from SLE patients contain low titers of autoantibodies to IFNα, as determined by a biosensor immunoassay (19). This autoantibody production could be due to the combination of elevated IFNα levels and B cell hyperactivity in patients with SLE. In the present study we detected low levels of anti-IFNα autoantibodies in the serum of 1 patient at baseline and in 2 of 7 placebo-treated patients at later time points. Therefore, in addition to breaking B cell tolerance to IFNα, IFN-K might induce anti-IFNα antibody production in some individuals by boosting or promoting the maturation of preexisting autoantibodies.

In patients with a positive IFN signature, kinoid administration significantly neutralized IFNα-mediated overexpressed genes, sometimes down to levels found in healthy individuals. In all patients, the magnitude of the inhibition of the IFN signature, as quantified by the IFN score, corresponded to the anti-IFNα titers. The link between anti-IFNα antibody titers and in vivo neutralization of the IFN signature is a strong indication that the IFN-K–induced variations in gene expression were related to true effects of the drug rather than random or confounding events. Moreover, patients with a higher IFN score at baseline produced higher anti-IFNα antibody titers in response to IFN-K, corresponding to a greater decrease in IFN score. These correlations suggest that IFN-K might be most effective in SLE patients who need it the most.

Morimoto et al (19) showed that most SLE patients with low levels of spontaneous anti-IFNα autoantibodies have not only lower type I IFN bioactivity, but also reduced disease activity. Also, in a clinical trial, sifalimumab, a passively administered IFNα-blocking monoclonal antibody, tended to reduce the number of disease flares and the need for immunosuppressive drugs (8). Coupled with the correlation between IFN-inducible gene expression and disease activity, the ability of IFN-K to reduce IFN scores suggests that it will also reduce disease activity itself. In fact, we found that the level of anti-IFNα antibody production corresponded to increased complement C3 levels, supporting the notion of a preclinical effect in this regard. We did not observe a significant effect on anti-dsDNA antibody levels, but their titers were low at baseline in most of the IFN-K–treated patients. We also did not find a significant effect on clinical outcomes. However, the lack of a clinical effect, especially on the SLEDAI score and the BILAG index, is not surprising given the small size and short duration of the study, the number of different dosages used, the nature and variability of the measures used, and the possibility that IFN-K effects were masked by corticosteroids in some patients.

Taken together, these initial results are promising and indicate that further clinical studies of the safety and efficacy of IFN-K in SLE are warranted. The observation that IFN-K activity correlates with baseline expression of IFN-induced genes opens clinically relevant perspectives for future selection of patients, based on clinical and biologic indices of disease activity. From a more general perspective, these data are the first demonstration of the feasibility of a novel type of therapy for autoimmune disorders, based on active immunization against cytokines produced in excess.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Vandepapelière had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Lauwerys, Grouard-Vogel, Fanget, Dhellin, Vandepapelière, Houssiau.

Acquisition of data. Hachulla, Spertini, Lazaro, Jorgensen, Mariette, Haelterman, Grouard-Vogel, Vandepapelière, Houssiau.

Analysis and interpretation of data. Lauwerys, Mariette, Grouard-Vogel, Vandepapelière, Houssiau.

ROLE OF THE STUDY SPONSOR

Neovacs SA had no role in the study design or in the collection, analysis, or interpretation of the data, the writing of the manuscript, or the decision to submit the manuscript for publication, other than the roles indicated in the Author Contributions section. Publication of this article was not contingent upon approval by Neovacs SA.

Acknowledgements

The authors thank Sarah Créac'h (Neovacs SA) for global study management and Dr. Phillip Leventhal (4Clinics, Paris, France) for medical writing assistance. Medical writing assistance was paid for by Neovacs SA. The authors thank Dr. Rasho Rashkov, MHAT, “Sveti Ivan Rilski,” Sofia, Bulgaria, for including patients in the clinical study. The authors also thank Dr. Zahir Amoura, Department of Internal Medicine, Assistance Publique-Hôpitaux de Paris, Pitié-Salpêtrière Hospital, Paris, France, for screening patients, although they were not enrolled in the study.

Ancillary