Safety and pharmacodynamics of rontalizumab in patients with systemic lupus erythematosus: Results of a phase I, placebo-controlled, double-blind, dose-escalation study

Authors

  • Jacqueline M. McBride,

    Corresponding author
    1. Genentech, South San Francisco, California
    • Genentech Inc., 1 DNA Way, South San Francisco, CA 94080
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Jenny Jiang,

    1. Genentech, South San Francisco, California
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Alexander R. Abbas,

    1. Genentech, South San Francisco, California
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Alyssa Morimoto,

    1. Genentech, South San Francisco, California
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Jing Li,

    1. Genentech, South San Francisco, California
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Romeo Maciuca,

    1. Genentech, South San Francisco, California
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Michael Townsend,

    1. Genentech, South San Francisco, California
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Daniel J. Wallace,

    1. Cedars-Sinai Medical Center and David Geffen School of Medicine, University of California, Los Angeles
    Search for more papers by this author
    • Dr. Wallace has received consulting fees from Genentech, UCB, Human Genome Sciences, GlaxoSmithKline, Teva, Cephalon, Pfizer, and Amgen (less than $10,000 each).

  • William P. Kennedy,

    1. Genentech, South San Francisco, California
    Search for more papers by this author
    • Dr. McBride, Ms Jiang, and Drs. Abbas, Morimoto, Li, Maciuca, Townsend, and Kennedy may own stock or stock options in F. Hoffmann-La Roche Ltd., the parent company of Genentech, Inc., as part of their general employment compensation.

  • Jorn Drappa

    1. Genentech, South San Francisco, California
    Search for more papers by this author

  • ClinicalTrials.gov identifier: NCT00541749.

Abstract

Objective

Systemic lupus erythematosus (SLE) is a chronic autoimmune disease characterized by the presence of autoantibodies and inflammation in multiple organ systems. Elevation of messenger RNA levels of interferon (IFN)–regulated genes (IRGs) has been described in the peripheral blood of SLE patients and has been associated with disease activity. The safety, tolerability, pharmacokinetics (PK), and pharmacodynamics (PD) of rontalizumab, a humanized IgG1 monoclonal antibody that neutralizes IFNα, were assessed in a phase I dose-escalation study of single and repeat doses of rontalizumab in adults with mildly active SLE. The present report describes the safety results and the impact of rontalizumab on expression of IRGs, IFN-inducible proteins, and autoantibodies.

Methods

Patients were enrolled into dose groups ranging from 0.3 to 10 mg/kg, administered via intravenous (IV) or subcutaneous routes. Expression levels of 7 IRGs and IFN-inducible serum proteins were monitored as potential biomarkers for the PD activity of rontalizumab.

Results

An acceptable safety profile was demonstrated for rontalizumab in patients with SLE. Prespecified criteria for dose-limiting toxicity were not met. The incidence of serious adverse events was comparable across cohorts. The PK properties were as expected for an IgG1 monoclonal antibody and were proportional to dose. Following administration of rontalizumab, a rapid decline in the expression of IRGs was observed in the 3 mg/kg and 10 mg/kg IV cohorts, and this effect could be sustained with repeat dosing. There was no apparent decline in the levels of IFN-inducible proteins or levels of anti–double-stranded DNA and anti–extractable nuclear antigen autoantibodies following treatment with rontalizumab.

Conclusion

The preliminary safety, PK profile, and observed PD effects of rontalizumab support further evaluation of its safety and efficacy in SLE.

Systemic lupus erythematosus (SLE) is a chronic autoimmune disease characterized by the presence of various autoantibodies, many of which bind to nuclear components such as nucleic acids and nucleic acid binding proteins. It is thought that dysregulated interferon-α (IFNα) signaling may have a causal role in disease pathogenesis. Elevated levels of type I IFNs have been observed in SLE sera (1–4), and an increase in the expression of IFN-regulated genes (IRGs), as well as polymorphisms in genes within the IFN pathway, has been described by several groups (5–14). The presence of the IFN signature has been shown to correlate with the presence of antibodies against extractable nuclear antigens (anti-ENAs) and anti–double-stranded DNA (anti-dsDNA) antibodies (2, 8), as well as with clinical manifestations (2, 15, 16). Moreover, development of SLE-like syndromes in some patients treated with recombinant IFNα, in the case of chronic viral infection or malignancy, further support this hypothesis of a role for IFNα in disease pathogenesis (17–22).

Rontalizumab is a humanized IgG1 monoclonal antibody that is designed to bind and neutralize all known subtypes of human IFNα and is being developed as a potential therapeutic agent for the treatment of SLE. In the present report, we describe our findings from a completed phase I, placebo-controlled, double-blind, dose-escalation study of rontalizumab in patients with mildly active SLE, the first rontalizumab trial in humans. The primary objective was to evaluate the safety and tolerability of rontalizumab, and a secondary objective was to characterize the pharmacokinetics (PK) profile. An important exploratory objective was to evaluate the pharmacodynamic (PD) activity of rontalizumab in order to demonstrate its pharmacologic activity and gain insight into the relationship between PD activity and drug exposure.

For our studies, we first conducted microarray analyses to establish the presence of the IFN signature within this specific patient population. The elevated IFN signature comprising the IFN-regulated genes (IRGs) was successfully identified in the patients at baseline, despite the fact that this patient population had only mildly active SLE. We then monitored the expression levels of selected IRGs derived from the larger IFN signature throughout the trial, using quantitative reverse transcription–polymerase chain reaction (qPCR). In addition, levels of selected circulating IFN-inducible proteins and autoantibodies were measured in the serum, both to assess their potential utility as protein biomarkers for the PD activity of rontalizumab and to identify possible correlations with IRG expression levels.

PATIENTS AND METHODS

Patients.

All patients were between the ages of 18 and 65 years and met the American College of Rheumatology diagnostic criteria for SLE (23). Patients had documented antinuclear antibodies (minimum titer 1:80), anti-dsDNA antibodies, and/or anti-Sm antibodies. Patients with organ-specific or life-threatening manifestations of SLE and those taking immunosuppressive medications or steroids exceeding a dose of 20 mg/day prednisone or equivalent were excluded. Other exclusion criteria included severe infections and chronic viral infections, such as frequent episodes of herpes (types 1 and 2). Patients provided informed consent, and the trial was conducted in accordance with the guidelines set forth by the International Conference on Harmonization.

Study objectives and design.

The primary objective for this trial was evaluation of the safety and tolerability of rontalizumab in patients with stable, mildly active SLE. Specific safety concerns were acute hypersensitivity reactions and infection, with a focus on viral reactivation. Additional key objectives included evaluation of the PK profile and PD effects of single- and repeat-dose administration of rontalizumab, as outlined in the study schema (Figure 1). Patients were permitted to continue taking stable concomitant medications, including nonsteroidal antiinflammatory drugs, antimalarials, and steroids at a dosage of up to 20 mg/day prednisone or equivalent. Patients who required steroid doses exceeding 0.5 mg/kg prednisone or equivalent or immunosuppressive regimens were discontinued from the study treatment and were entered into the 48-week safety followup period.

Figure 1.

Study schema and design. A total of 60 patients were enrolled in 6 cohorts at a ratio of 4:1 for active treatment to placebo. Each cohort included 8 patients administered rontalizumab and 2 patients administered placebo. The dose levels of rontalizumab tested were 0.3 mg/kg intravenously (IV), 1 mg/kg IV, 3 mg/kg IV, 10 mg/kg IV, 1 mg/kg subcutaneously (SC), and 3 mg/kg SC. Enrollment to the next-higher dose level began once patients receiving a given dose had been observed for at least 14 days and the prespecified dose-limiting criteria were not met. In the first stage, patients received a single dose followed by a minimum 10-week washout period. Patients returned to enter the repeat-dose stage and were administered 3 repeat doses (once per month, on days 0, 28, and 56) at the same dose and route of administration as that in the single-dose stage. Patients were followed up until day 252 after receiving the first repeat dose.

Samples collected from the trial.

Whole blood samples were collected in Qiagen PAXgene blood RNA tubes from healthy subjects and patients at screening, on day 0 (predose), and on days 3, 7, 14, 28, and 56 after single-dose administration of rontalizumab, as well as on day 0 (predose) and on days 28, 56, 63, 84, 112, 168, 224, and 252 during the repeat-dose stage. Serum samples were also collected from healthy subjects (Bioreclamation) as well as from the SLE patients in the trial at the time points described.

Total RNA processing and complementary DNA (cDNA) generation.

RNA was extracted from the whole blood samples using the Qiagen PAXgene Blood RNA kit. The RNA quality was measured using an Agilent 2200 Bioanalyzer, and concentrations of RNA were determined using a Nanodrop 2000 (Thermo Scientific). Following processing of the RNA, cDNA was generated using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems).

Quantitative PCR with low-density arrays.

For qPCR analyses, we used TaqMan Low Density Array cards (Applied Biosystems) containing cDNA primers/probes for the following genes: IFI27, IFI44, IFIT1, MX1, OAS1, OAS2, and OAS3, with GAPDH as the calibrator. The cDNA reactions were prepared with TaqMan Universal PCR Master Mix (Applied Biosystems) and loaded onto the cards in accordance with the manufacturer's instructions. Data analysis was performed using SDS (version 2.2) and RQ Manager software (version 1.2; Applied Biosystems). IRG expression levels in a given sample, determined as threshold cycle (Ct) values, were first normalized to the values for a standard housekeeping gene, GAPDH (messenger RNA levels of GAPDH are unchanged in patient samples), to obtain the ΔCt for individual genes. The mean interferon signature metric (ISM) scores were calculated by taking the gene expression values (ΔCt) for the 7 selected IRGs at baseline and multiplying by −1 to show the correct directionality of relative expression (mean − ΔCt).

PK analyses.

Serum levels of rontalizumab were evaluated before dosing and at 1 hour and 4 hours after dosing, and then on days 1, 3, 7, 10, 14, 21, 28, 42, and 56 after single-dose administration. PK parameters of rontalizumab were determined using a noncompartmental method (WinNonlin; Pharsight). The area under the concentration–time curve was calculated by the linear trapezoidal method.

Gene expression microarray analysis.

Labeled cDNA was generated from RNA derived from the whole blood of patients in the placebo, 3 mg/kg intravenous (IV) rontalizumab, and 10 mg/kg IV rontalizumab groups in the single-dose stage. Clustering results were visualized using Java Treeview version 1.08 (provided by M. Eisen, Berkeley, CA). Differential expression of genes was assessed by linear modeling via the Limma package (Bioconductor). Moderated t-statistics from modeling were used to calculate the adjusted P values, using the Benjamini-Hochberg correction (24). Differential expression was defined as cases in which the fold change in gene expression was greater than 1.5 and the adjusted P value for the difference in gene expression was less than 0.2 (additional details are found in the Supplementary Methods, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131).

Serum IFN-inducible protein assessment.

Meso Scale Discovery plates were used to measure the IFN-inducible T cell α chemoattractant (I-TAC), monocyte chemoattractant protein 1 (MCP-1), and B lymphocyte chemoattractant (BLC) proteins in serum samples. IFNγ–inducible protein 10 (IP-10) was measured using commercially available kits (R&D Systems). BAFF protein levels were measured using an in-house–generated immunoassay.

Serum autoantibody assessment.

Antibodies against the autoantigens Sm, RNP, Ro (anti–Ro/SSA 60 and anti–Ro/SSA 52), La, Scl-70, Jo-1, ribosomal P, and chromatin were measured in serum samples using a Quantaplex SLE Profile 8 assay (Inova Diagnostics). Serum levels of antibodies specific to dsDNA were measured using a Human Anti-dsDNA Enzyme Immunoassay kit (Inova Diagnostics).

RESULTS

Demographics and baseline characteristics of the patients.

For this trial, 60 patients were enrolled at ∼20 centers in the United States. Most patients (95%) were female, with a mean age of 47 years. Among the patients, 70% were Caucasian, and 27% were African American. The mean disease duration was ∼9 years, and the mean ± SD score for the extent of disease activity at baseline, as measured by the Safety of Estrogens in Lupus Erythematosus National Assessment version of the Systemic Lupus Erythematosus Disease Activity Index (25), was 3.4 ± 2.7. Only 30% of the patients were receiving prednisone at baseline (mean dosage 7 mg/day).

Safety results.

Phase I of the rontalizumab study was designed to evaluate the preliminary safety of a single dose or multiple doses of rontalizumab, administered IV or subcutaneously (SC), with limited patient exposures. The prespecified dose-limiting toxicity criteria were not met when rontalizumab was administered in doses ranging up to 10 mg/kg. Both IV and SC administration of rontalizumab was generally well tolerated, with no reported grade 3 infusion/injection site reactions. None of the adverse events (AEs) that occurred led to the discontinuation of study drug administration.

Most of the reported AEs were mild or moderate (grade 1 or grade 2 according to version 4 of the National Cancer Institute's Common Terminology Criteria for Adverse Events). The 5 most commonly reported AEs were upper respiratory infections, nausea and vomiting, headaches, musculoskeletal and connective tissue signs and symptoms, and urinary tract infections. The rates and incidence of infection-related AEs were similar across dose groups (Table 1).

Table 1. Summary of adverse events (AEs) across dose groups in patients with systemic lupus erythematosus receiving either placebo or rontalizumab, administered either intravenously (IV) or subcutaneously (SC)*
 Placebo (n = 12)Rontalizumab
0.3 mg/kg IV (n = 8)1 mg/kg IV (n = 8)1 mg/kg SC (n = 8)3 mg/kg IV (n = 8)3 mg/kg SC (n = 8)10 mg/kg IV (n = 8)Total (n = 48)
  • *

    Except where indicated otherwise, values are the number (%) of patients. SAE = serious adverse event; 95% CI = 95% confidence interval.

≥1 AE11 (91.7)8 (100)8 (100)8 (100)8 (100)7 (87.5)8 (100)47 (97.9)
 Grade 2 or higher6 (50)8 (100)5 (62.5)6 (75)6 (75)5 (62.5)5 (62.5)35 (72.9)
 Grade 3 or higher1 (8.3)2 (25)1 (12.5)3 (37.5)2 (25)2 (25)2 (25)12 (25)
≥1 SAE1 (8.3)1 (12.5)02 (25)2 (25)2 (25)07 (14.6)
Drug-related AE        
 ≥1 drug-related AE within 24 hours1 (8.3)01 (12.5)3 (37.5)1 (12.5)3 (37.5)2 (25)10 (20.8)
 ≥1 drug-related AE2 (16.7)1 (12.5)5 (62.5)4 (50)4 (50)3 (37.5)3 (37.5)20 (41.7)
Infection        
 ≥1 infection AE9 (75)7 (87.5)7 (87.5)7 (87.5)7 (87.5)5 (62.5)6 (75)39 (81.3)
 Rate of infections per patient-year (95% CI)1.25 (0.9–1.8)0.73 (0.4–1.3)0.96 (0.6–1.6)1.45 (0.9–2.3)1.91 (1.3–2.8)1.15 (0.7–2.0)1.72 (1.1–2.6)1.28 (1.1–1.5)
 ≥1 infection SAE00001 (12.5)001 (2.1)

There was some imbalance in the reported frequencies of AEs that were grade 3 or higher and in the reported frequencies of serious adverse events (SAEs) between the SLE patients receiving placebo and SLE patients receiving rontalizumab. In total, 1 patient (8.3%) in the placebo group and 12 patients (25%) in the combined rontalizumab group reported experiencing AEs of grade 3 or higher. Despite this imbalance, there was no overall dose–AE relationship for grade 3 or higher AEs when patients were treated with rontalizumab at doses ranging from 0.3 mg/kg to 10 mg/kg.

Similarly, the proportion of patients who experienced an SAE (total of 10 SAEs reported) was 8.3% (n = 1) in the placebo group and 14.6% (n = 7) in the combined rontalizumab group (Table 1). The SAEs included a single case of each of the following events: musculoskeletal chest pain, abdominal pain of unknown cause, appendicitis, high-grade coronary artery stenosis, cerebral atherosclerosis, atrial fibrillation, and appendicitis. Each of these SAEs was classified by the investigators as unrelated to the study drug.

Infections.

Almost all of the reported infections were mild or moderate (grade 1 or grade 2 infection AEs), except for 2 infections that were grade 3 (sinusitis and bronchitis), occurring in the 10 mg/kg IV rontalizumab cohort. Certain infections, such as urinary tract infections, herpes reactivation, and unspecified viral infections, occurred at a higher rate in the combined rontalizumab group than in the placebo group. One serious infection (a case of appendicitis) was reported in the 3 mg/kg IV rontalizumab dose group.

Malignancy.

One case of malignancy was reported in the 3 mg/kg SC rontalizumab cohort. This was a case of acute myelogenous leukemia in a 44-year-old woman with a 17-year history of SLE. This patient presented with decreasing white cell counts shortly after completion of the repeat-dose phase of the trial and 338 days after her first dose of rontalizumab. Further evaluation resulted in a diagnosis of acute myelogenous leukemia. No clear risk factors, such as prior exposure to alkylating agents, were present.

PK properties of rontalizumab.

Overall, the PK properties of rontalizumab were found to be dose proportional within the dose range tested, when rontalizumab was administered either IV or SC as a single dose or in multiple doses. The clearance rate of rontalizumab was 2.88 ml/day/kg and the terminal half-life was 18.8 days. The bioavailability following SC administration of rontalizumab was ∼44% (additional PK parameters are shown in Supplementary Table 1, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131).

Evidence of the IFN signature in 50% of patients at baseline.

Whole blood–derived RNA from patients in the placebo (n = 12), 3 mg/kg IV rontalizumab (n = 8), and 10 mg/kg IV rontalizumab (n = 8) groups was assayed in gene expression microarrays. Genome-wide hierarchical clustering of the gene expression data obtained from the samples at baseline revealed a major subset of SLE patients expressing IRGs in greater abundance when compared to the expression profiles in the remaining patient population. The main branches of the dendrogram, as shown in Figure 2A, indicate that this subset with elevated IRG expression levels represented ∼50% of the patients.

Figure 2.

Identification of interferon (IFN)–regulated genes (IRGs) as evidence of the IFN signature in patients with mildly active systemic lupus erythematosus (SLE). A, Microarray analyses of the top 20% most-variable genes in patients with SLE at baseline identified IFN signature genes in the main cluster. The relative abundance of IRG expression in patients was categorized as high (Hi) versus low (Lo). The IFN signature genes as defined by Baechler et al (11) were identified in the main cluster, and the density of their incidence was plotted in the blue box to confirm that their distribution was congruent with the peak signature. B, A bimodal distribution of the mean IFN signature metric (ISM) scores for the 7 selected IRGs was observed in SLE patients compared to healthy controls. ISM scores were derived from quantitative polymerase chain reaction (qPCR) analyses of differential IRG expression, using a threshold of −1.5 (horizontal broken line) to categorize ISMhigh and ISMlow patients. C, Hierarchical clustering analyses identified the top 46 genes that were the most differentially expressed between ISMhigh and ISMlow patients.

The specific genes used to define our IFN signature gene set were identified from this cluster by overlaying a set of previously published IRGs (11) onto the gene dimension of the heatmap. The point of greatest density was selected as the core of the signature set, and the set of genes was then expanded from that point to include all genes that were correlated at a coefficient of greater than 0.5. The size of the signature was moderate, with 96 genes from the region showing a correlation of gene expression levels at a Spearman's correlation coefficient of >0.5.

Determination of the ISM score.

Quantitative PCR has a significantly broader dynamic range than microarray analysis. Therefore, qPCR assays were developed to monitor the relative expression levels of a representative set of IRGs, including IFI27, IFI44, MX1, IFIT1, OAS1, OAS2, and OAS3 (9, 11). Additional information with regard to the choice of these genes is provided in the Supplementary Methods (available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131).

To determine the ISM score, the average expression level of 7 selected IRGs (mean −ΔCt) was determined at baseline in samples from each patient and also from samples derived from normal healthy subjects. All patients and healthy subjects had abundant constitutive gene expression. Density plots showing the distribution of the ISM scores demonstrated a clear bimodal distribution, whereby slightly more than 50% of the patients (31 of 60) exhibited ISM scores that overlapped with the levels found in healthy controls (Figure 2B). A conservative threshold of −1.5 (mean −ΔCt) was set to identify differential IRG expression falling either above or below this cutoff, and thus patients could be designated as either ISMhigh (29 [48%] of 60 patients) or ISMlow (31 [52%] of 60 patients), respectively.

The 7 IRGs selected for our analyses are representative of the entire IFN signature gene set. The ISM designation of patients was also used to determine whether the majority of the other IFN signature genes would similarly show differences in expression. Linear modeling of microarray expression data for known genes by this discrete ISMhigh or ISMlow classification, as well as filtering for genes that were elevated at least 1.5 fold in between-group comparisons and that had a false discovery rate of less than 20%, yielded a set of 46 differentially expressed IFN signature genes. Figure 2C shows the clustering of the microarray data from these samples obtained from patients at baseline (n = 28), listed in ascending order, with stratification according to the qPCR-derived ISM score. The expression profiles of these 46 genes were highly similar (mean ± SD Spearman's rho = 0.787 ± 0.135) to the qPCR-derived ISM score, with the exception of that in 1 patient (Spearman's rho = 0.672 ± 0.098). These data confirm that the 7 selected IRGs are representative of the IFN signature identified in our microarray analyses.

Dose-dependent reduction in levels of IRGs with single and repeat doses of rontalizumab.

IRG expression was evaluated longitudinally by qPCR using all patient samples at baseline (predose) and at the visits following single and repeat-dose administration of rontalizumab. The median expression levels of a representative gene, IFI27, after treatment relative to that at baseline (calculated as 2math image × 100) in all patients within each IV rontalizumab dose group and the placebo group are shown in Figure 3. All 7 of the IRGs analyzed demonstrated a similar pattern of response to rontalizumab (results not shown).

Figure 3.

Dose-dependent reduction in interferon-regulated gene (IRG) expression. Expression levels of IRGs after treatment relative to those at baseline (calculated as 2math image × 100) were determined in each treatment group. The Ct values were first normalized to those for GAPDH (ΔCt), and changes in gene expression were calculated relative to the levels at baseline (ΔΔCt). The median relative change in expression of a representative IRG, IFI27, is shown for each intravenous (IV) rontalizumab dose group and the placebo group, plotted against the number of days postadministration. Plots for both the single-dose stage (left) and the repeat-dose stage (right) are shown. Arrows indicate the days of dosing.

Rontalizumab administration at a dose of 3 mg/kg IV and 10 mg/kg IV induced a substantial decline (>50%) in relative gene expression, compared to baseline values, in all of the patients who had received a single and/or repeat dose. The median expression levels within each dose group reached 31.3% of baseline in the 3 mg/kg IV group and as low as 23.1% of baseline in the 10 mg/kg IV group, after single-dose administration (Figure 3, left). In the 10 mg/kg IV dose group, the maximum suppression achieved in individual patients following administration of a single dose of rontalizumab ranged from 2.1% to 55.5% of baseline and could be sustained beyond 28 days after dosing. During the washout period, the majority of patients demonstrated a recovery in IRG expression to levels present at baseline.

Repeat dosing of rontalizumab in the 3 mg/kg IV and 10 mg/kg IV groups induced a dose-dependent decline in IRG levels to an extent similar to that measured in the single-dose stage, and the decline was sustained with repeat dosing (Figure 3, right). Moreover, all patients in the 10 mg/kg IV dose group maintained this decline in IRG expression for more than 100 days after receiving the last repeat dose (ranging from 18.6% to 29.9% of baseline). In contrast, at the lowest IV dose of 0.3 mg/kg, there was little change following either single or repeat dosing. At the 1 mg/kg IV dose level, repeat-dose administration of rontalizumab appeared to result in an increased magnitude and duration of suppression of IRG expression when compared to that after single-dose administration. However, the extent of this suppression was still not comparable to that measured in the higher-dose IV groups.

Six months after the last of the repeat doses was administered, there was evidence of recovery in IRG expression in all patients. Patients within the SC dose groups showed significant variability in the response to rontalizumab and only a modest decline in gene expression with repeat-dose administration at the doses tested. The bioavailability was such that the exposures achieved in patients receiving the 1 mg/kg SC and 3 mg/kg SC doses were only marginally higher than that observed in the lowest dose IV cohorts. All patients, irrespective of the baseline ISM score (ISMhigh or ISMlow), showed a similar magnitude and duration of decline in IRG expression (additional data are shown in Supplementary Figure 1, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131).

Elevation of IFN-inducible proteins in ISMhigh patients.

To evaluate the correlation between IRG expression in whole blood and expression of IFN-inducible proteins in the serum, IFN-inducible proteins, including IP-10, I-TAC, BLC (CXCL13), MCP-1, and BAFF, were measured. These proteins have been reported to be elevated in ISMhigh SLE patients when compared to ISMlow SLE patients (26). Patients were separated by ISM category, and serum protein levels at baseline were measured and compared to levels found in the serum of normal healthy subjects (Figure 4). It was evident that circulating serum BLC and I-TAC levels in ISMhigh patients were significantly elevated compared to those in ISMlow patients at baseline (P < 0.05). BLC, MCP-1, and I-TAC levels were also significantly elevated (P < 0.05) in both ISMhigh and ISMlow patients compared to the levels measured in the serum from normal healthy subjects (Figure 4). Serum levels of IP-10 or BAFF, although abundant, did not appear different between the SLE patients in this study and healthy controls. Serum levels of these proteins did not change following a single dose (Figure 4) or repeat doses (results not shown) of rontalizumab.

Figure 4.

Baseline levels (left panels) and longitudinal measurements (right panels) of interferon (IFN)–inducible proteins. Left panels, Predose levels of B lymphocyte chemoattractant (BLC), monocyte chemoattractant protein 1 (MCP-1), IFN-inducible T cell α chemoattractant (I-TAC), BAFF, and IFNγ-inducible protein 10 (IP-10) were measured in the serum of patients with systemic lupus erythematosus (n = 28) (separated by IFN signature metric [ISM] category of high [Hi] versus low [Lo] IFN-regulated gene expression), in comparison with healthy controls (n = 20). Each open circle represents an individual subject. Results are shown as box plots, where each box represents the 25th to 75th percentiles, lines inside the boxes represent the median, and lines outside the boxes represent the 10th and 90th percentiles. Right panels, Levels of each protein were monitored longitudinally in patients after administration of placebo, 3 mg/kg intravenous (IV) rontalizumab, or 10 mg/kg IV rontalizumab. Bars show the median with 25th to 75th percentiles, from the single-dose stage only. ∗ = significant difference (P < 0.05), as determined by pairwise comparisons using the nonparametric Wilcoxon test.

Elevation of autoantibodies in ISMhigh patients.

Although patients had stable, mildly active disease, many subjects had measurable levels of circulating autoantibodies at baseline, including antibodies against dsDNA and ENAs. Moderate-to-high levels of anti–Ro/SSA 60, anti–Ro/SSA 52, and anti-RNP antibodies were detected in several patients, and levels were significantly higher (P < 0.05) in ISMhigh patients compared to ISMlow patients (Figure 5A). Longitudinally, autoantibody levels in the serum samples obtained from patients receiving placebo and patients receiving either 3 mg/kg or 10 mg/kg IV rontalizumab were relatively stable. Of interest, these serum autoantibody levels were unaltered by rontalizumab treatment following administration of either a single dose (Figure 5B) or repeat dose (results not shown).

Figure 5.

Elevated levels of anti–double-stranded DNA (anti-dsDNA), anti-RNP, and anti-Ro/SSA (60 kd and 52 kd) antibodies in the serum of patients with systemic lupus erythematosus with high expression of interferon (IFN)–regulated genes (IRGs). A, SLE patients (n = 60) were separated by the defined IFN signature metric (ISM) score (high [Hi] versus low [Lo] IRG expression), and serum levels of anti-RNP and anti-Ro/SSA antibodies (expressed in light units [LU]) and absolute levels of anti-dsDNA antibodies (expressed in units) were determined. The broken horizontal line represents the limit of detection for each assay. Each open circle represents an individual patient. Results are shown as box plots, where each box represents the 25th to 75th percentiles, lines inside the boxes represent the median, and lines outside the boxes represent the 10th and 90th percentiles. ∗ = significant difference (P < 0.05), as determined by pairwise comparisons using the Wilcoxon test. B, Longitudinal measurements of each autoantibody were conducted in the serum of patients receiving placebo (n = 12), patients receiving 3 mg/kg intravenous (IV) rontalizumab (n = 8), and patients receiving 10 mg/kg IV rontalizumab (n = 8). Results are from individual patients in the single-dose stage, and are the results from only those patients with levels above the manufacturer's determined thresholds for a positive value. The total number of patients tested is given for each antibody measurement.

DISCUSSION

Evidential data suggest that dysregulated signaling through the type I IFN pathway plays a critical role in the pathogenesis of SLE. Increased signaling through this pathway leads to multiple biologic events that may initiate or sustain chronic inflammation and autoimmunity in this disease. These include increased activity and accelerated maturation of antigen-presenting cells, increased B cell maturation and autoantibody production, and T cell activation. Therefore, IFNα is an attractive target for therapeutic intervention (27–29).

This report summarizes the results of a recently completed phase I, first-in-human trial of rontalizumab, a monoclonal antibody that neutralizes all IFNα subtypes that occur in humans. The trial's primary objective was to characterize the safety and tolerability of rontalizumab in patients with stable, mildly active SLE. This population was chosen as the patient cohort for preliminary safety evaluation of a potentially immunosuppressant agent to minimize the confounding influence of coadministered medications, including high-dose steroids and other immunosuppressive agents. During the trial, treatment with immunosuppressive agents was not permitted, and the maximum dosage of steroids allowed was 20 mg/day of prednisone or equivalent.

Overall, rontalizumab was generally safe and well tolerated in this phase I trial in patients with mildly active SLE. In light of the important role of type I IFN in host defense, an important question was whether rontalizumab would increase the risk of infections, particularly viral infections. As shown in Table 1, the exposure-adjusted rate of infections was similar between treatment groups, and no dose-related increase in infections was observed. Although there was an imbalance in the SAE frequency in the rontalizumab cohorts compared to the placebo cohort, none of the reported SAEs were attributed to drug treatment by the study investigators. In addition, these AE and SAE frequencies are based on a small number of events. There was also an imbalance in the frequency of AEs that were grade 3 or higher in the rontalizumab-treated SLE patients compared to SLE patients receiving placebo. Future clinical studies will require continued safety monitoring of rontalizumab in SLE patients.

One case of malignancy was reported in a patient in the 3 mg/kg SC cohort. This was a case of acute myelogenous leukemia in a 44-year-old woman with a 17-year history of SLE. There is evidence to suggest that the type I IFN pathway may be involved in tumor surveillance, and mice with a complete absence of type I IFN signaling due to a targeted deletion of the IFNAR gene have been reported to show accelerated growth of implanted tumors or chemically induced cancers (30–33). Whether selective pharmacologic neutralization of IFNα activity that spares other type I IFN activity will increase the risk for malignancy is unknown. Continued monitoring for malignancy and exclusion of SLE patients with high risk for malignancy in future studies will be necessary to understand the potential risk from selective IFNα neutralization.

Another important objective of this trial was to measure the PD effects of rontalizumab, specifically its ability to reduce the overexpression of IRGs, which is a characteristic observed in many patients with SLE. This information was critical for selecting appropriate doses for further study in phase II. Since measuring IFNα activity or protein levels is technically challenging, measuring the expression levels of genes directly downstream of the IFN pathway was the approach selected.

As a first step, we confirmed the presence of the IFN signature in our specific trial population, since prior studies had focused on patients with moderate-to-severe disease activity, whereas patients in this trial had mild disease. Microarray analyses revealed the presence of elevated IRG expression in almost one-half of the patients, and the genes within the signature identified in this study were consistent with previously described IFN signature genes (9, 11). To track changes in expression following exposure to rontalizumab, genes that were representative of the entire IFN signature were selected for qPCR analyses. Similar to the findings in the study by Hua et al (8), our data supported our choice of genes as being driven primarily by type I IFNs, such as IFNα, in SLE serum, with little contribution from IFNγ.

Analysis of IRG expression levels following administration of rontalizumab demonstrated a substantial dose-dependent decline in the expression of all selected genes (IFI27, MX1, IFI44, IFIT1, OAS1, OAS2, and OAS3). At both the 3 mg/kg and 10 mg/kg IV doses, this reduction was evident in the majority of patients and could be sustained beyond 28 days after dosing. At the lower doses of 0.3 mg/kg and 1 mg/kg IV, the decline was short-lived and similar to the biologic variability found in the patients receiving placebo. The PD effect of rontalizumab was reversible at all dose levels, and IRG expression levels tended to revert to baseline levels during the 10-week washout period. Following the administration of 3 doses of rontalizumab, administered once per month, the extent of decline in IRG expression in the 3 mg/kg and 10 mg/kg IV groups was very similar to that observed after a single dose but was sustained for more than 3 months after the last repeat dose. There were only modest and transient changes in IRG expression within the SC dose groups, which may be explained by the exposure provided by the SC route of administration, which has a bioavailability of ∼44%.

The decline in IRG expression following treatment with the higher IV doses was similar in all patients regardless of their ISM designation at baseline. None of the patients who were designated as being ISMhigh reached the levels of IRG expression seen in the patients designated as being ISMlow nor did they reach the basal levels detected in healthy individuals. This result is likely attributable to the residual activity of other type I IFNs, such as IFNω, IFNβ, or other drivers that are also present in the serum of these patients and have the capacity to induce expression of the same genes. It is not currently known whether complete or partial normalization of IRG expression levels is required in order to achieve efficacy.

Importantly, this study was not designed to study the effect of rontalizumab on clinical measures of disease activity. The trial population had mildly active disease at baseline. Therefore, the relationship between IRG expression at baseline and a rontalizumab treatment effect remains to be established in larger trials involving SLE patients with greater disease activity.

Additional protein analytes were also assessed to evaluate their utility, if any, as PD biomarkers and to determine whether the levels of these proteins were correlated with the changes in IRG expression. It has been reported that some IFN-inducible proteins, such as I-TAC, IP-10, MCP-1, and BLC (CXCL13), were elevated in the serum of patients with SLE bearing a higher ISM score (26, 34). We reproduced these findings in that we observed elevated levels of BLC, I-TAC, and MCP-1, but not BAFF or IP-10, in the SLE serum, which is promising, even within this population of patients with mild disease. However, the protein levels did not decline following rontalizumab administration, and thus it is likely that these proteins are also regulated by other inflammatory stimuli, including residual IFNs such as IFNω, IFNβ, IFNγ, and IFNλs, as well as tumor necrosis factor α, or even bacterial or viral products present in the circulation (35–37).

A relationship between the presence of IFN activity and certain autoantibodies against ENAs and anti-dsDNA antibodies has been described (8, 10, 15). Through our ranking of patients by ISM score, it was apparent that those patients with measurable levels of anti-Ro or anti-RNP antibodies tended to be found in the ISMhigh patient subset. There also appeared to be 2 distinct subsets of patients, those with both anti-Ro and anti-RNP antibodies and those with anti-RNP antibodies only. A similar trend was evident with respect to anti-dsDNA antibodies, in which patients with higher levels of anti-dsDNA were mainly concentrated in the ISMhigh subset. Autoantibody levels did not decline following administration of rontalizumab. It is likely that these autoantibodies are derived from long-lived plasma cells, and longer treatment periods may be required to observe a measurable impact on circulating antibody levels.

In conclusion, we found that rontalizumab was well tolerated in patients with SLE in this phase I trial and that its safety profile supported further testing in larger trials. Administration of rontalizumab resulted in the expected PD effect of inducing a dose-dependent decline in the expression of IRGs. Subsequent studies, including the ongoing phase II trial in patients with moderately or severely active disease, will further investigate the PK/PD relationship as well as the nature of the relationship between IRG expression and clinical markers of response to treatment with rontalizumab.

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. McBride 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. Morimoto, Maciuca, Wallace, Drappa.

Acquisition of data. McBride, Li, Wallace, Drappa.

Analysis and interpretation of data. McBride, Jiang, Abbas, Li, Maciuca, Townsend, Wallace, Kennedy, Drappa.

ROLE OF THE STUDY SPONSOR

Dr. Wallace was the lead investigator in this trial, which was sponsored by Genentech, and was involved in trial design, patient recruitment, data analysis, and discussion of the results. All authors, some of whom are employees of Genentech, independently collected the data, interpreted the results, and had the final decision to submit the manuscript for publication. Genentech reserved the right to review and comment on this publication.

Acknowledgements

We thank Kristen Wolslegel, Zhenling Yao, Joe Simpson, Brendan Bender, and Don Sinclair for their help with this study.

Ancillary