Neutralization of interferon-╨Ю┬▒ inducible genes and downstream effect in a phase I trial of an antiinterferon-╨Ю┬▒ monoclonal antibody in systemic lupus erythematosus.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 60, No. 6, June 2009, pp 1785–1796 DOI 10.1002/art.24557 © 2009, American College of Rheumatology Neutralization of Interferon-␣/␤–Inducible Genes and Downstream Effect in a Phase I Trial of an Anti–Interferon-␣ Monoclonal Antibody in Systemic Lupus Erythematosus Yihong Yao, Laura Richman, Brandon W. Higgs, Christopher A. Morehouse, Melissa de los Reyes, Philip Brohawn, Jianliang Zhang, Barbara White, Anthony J. Coyle, Peter A. Kiener, and Bahija Jallal overexpression of messenger RNA for BAFF, TNF␣, IL-10, IL-1␤, GM-CSF, and their respective inducible gene signatures in whole blood and/or skin lesions, we observed a general trend toward suppression of the expression of these genes and/or gene signatures upon treatment with anti-IFN␣ mAb. Conclusion. IFN␣/␤-inducible gene signatures in whole blood are effective pharmacodynamic biomarkers to evaluate anti-IFN␣ mAb therapy in SLE. Anti-IFN␣ mAb can neutralize overexpression of IFN␣/␤-inducible genes in whole blood and lesional skin from SLE patients and has profound effects on signaling pathways that may be downstream of IFN␣ in SLE. Objective. Type I interferons (IFNs) play an important role in the pathogenesis of systemic lupus erythematosus (SLE). This phase Ia trial was undertaken to evaluate the safety, pharmacokinetics, and immunogenicity of anti-IFN␣ monoclonal antibody (mAb) therapy in SLE. During the trial, we also examined whether overexpression of an IFN␣/␤-inducible gene signature in whole blood could serve as a pharmacodynamic biomarker to evaluate IFN␣ neutralization and investigated downstream effects of neutralizing IFN␣ on BAFF and other key signaling pathways, i.e., granulocyte–macrophage colony-stimulating factor (GM-CSF), interleukin-10 (IL-10), tumor necrosis factor ␣ (TNF␣), and IL-1␤, in SLE. Methods. Affymetrix Human Genome U133 Plus 2.0 microarrays were used to profile whole blood and lesional skin of patients receiving standard therapy for mild to moderate SLE. Selected IFN␣/␤-inducible proteins were analyzed by immunohistochemistry. Results. With the study treatment, we observed anti-IFN␣ mAb–specific and dose-dependent inhibition of overexpression of IFN␣/␤-inducible genes in whole blood and skin lesions from SLE patients, at both the transcript and the protein levels. In SLE patients with Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by severe immune system defects and production of autoantibodies that lead to inflammation and tissue damage (1). SLE symptoms range from a mild rash to life-threatening nephritis and central nervous system disease. Current SLE therapies are aimed at control of symptoms and do not address the underlying causes of disease, and may entail risk of serious adverse effects (2). Novel therapies that address disease pathogenesis more directly and with less toxicity are greatly needed. Type I interferons (IFNs) have been implicated in autoimmune diseases (3–7), including SLE (1), and of note, evidence from gene expression profiling studies implicates type I IFNs in SLE (8–11). An especially important research observation is that elevated levels of IFN␣ are observed in the serum of a subset of SLE patients (8,12–16). To treat SLE by lowering IFN␣ levels, we have developed a fully human IgG1 monoclonal antibody (mAb) that binds to a majority of the subtypes of human IFN␣ and inhibits IFN-mediated signaling. A singledose, double-blind, placebo-controlled phase Ia trial ClinicalTrials.gov identifier: NCT00299819. Yihong Yao, PhD, Laura Richman, DVM, PhD, Brandon W. Higgs, PhD, Christopher A. Morehouse, MS, Melissa de los Reyes, BS, Philip Brohawn, BS, MBA, Jianliang Zhang, PhD, Barbara White, MD, Anthony J. Coyle, PhD, Peter A. Kiener, DPhil, Bahija Jallal, PhD: MedImmune, Gaithersburg, Maryland. Drs. Yao, Richman, Higgs, Zhang, Coyle, Kiener, and Jallal, and Mr. Morehouse, Ms de los Reyes, and Mr. Brohawn own stock or stock options in MedImmune, which is a subsidiary of AstraZeneca International. Dr. White owns stock options in AstraZeneca. Address correspondence and reprint requests to Yihong Yao, PhD, One MedImmune Way, Gaithersburg, MD 20878. E-mail: YaoY@MedImmune.com. Submitted for publication March 24, 2008; accepted in revised form March 2, 2009. 1785 1786 (MI-CP126) was conducted to study anti-IFN␣ mAb treatment in patients with mild to moderate SLE with cutaneous involvement who were receiving standard-ofcare therapy (17). The primary objective of the MICP126 trial was to evaluate the safety and tolerability of intravenously administered anti-IFN␣ mAb, over a dose escalation range of 0.3–30 mg/kg, as compared with placebo in adult SLE patients. Other objectives, reported here, were to evaluate pharmacodynamic effects of anti-IFN␣ mAb and, additionally, to study the effects of IFN␣ neutralization on downstream signaling pathways in SLE. Pharmacodynamic biomarkers are needed in early-phase clinical trials to demonstrate that a potential therapeutic molecule has altered its intended target, based on the proposed mechanism of action, at clinically achievable concentrations (18). Ideal pharmacodynamic biomarkers are sensitive, easy to measure, and present in easily accessible tissue, and correlate with disease activity in target tissue (18). Whole blood provides an easily accessible surrogate tissue for monitoring drug pharmacodynamics, especially when pharmacokinetic parameters are routinely measured in the peripheral blood to circumvent difficulties in assessing drug concentration at the disease site. Free IFN␣ protein in the serum of SLE patients would be the most reasonable choice as a pharmacodynamic marker for evaluating anti-IFN␣ therapy in SLE. However, our internal studies, as well as studies by other investigators (8,12,13), have demonstrated that only a fraction of SLE patients have measurable IFN␣ protein in the serum. IFN␣-inducible genes, in contrast, are directly downstream of the drug target, are overexpressed in whole blood from the majority of SLE patients, and their expression can be quantitatively measured with microarray or TaqMan quantitative real-time polymerase chain reaction (PCR)–based assays (9–11,19). In previous work, we used microarray transcript profiling and TaqMan quantitative real-time PCR to demonstrate overexpression of messenger RNA (mRNA) for type I IFN family members and a large panel of IFN␣/␤-inducible genes in whole blood from SLE patients. We found that the IFN␣/␤ signaling pathway was the most highly activated signaling pathway in SLE whole blood (19). We defined algorithms for using cytokine-inducible gene signature scores to evaluate cytokine activity in whole blood and at disease sites in inflammatory and autoimmune diseases (19,20). We also developed a panel of potential pharmacodynamic biomarkers for anti-IFN␣ mAb, comprising 21 IFN␣/␤inducible genes that are overexpressed in whole blood from SLE patients (19). YAO ET AL In this study, we tested this 21–IFN␣/␤-inducible gene signature as a pharmacodynamic biomarker for determining whether anti-IFN␣ mAb neutralizes IFN␣ in a specific and dose-dependent manner. In addition, we investigated the effects of anti-IFN␣ mAb on signaling pathways for interleukin-10 (IL-10) (21), granulocyte– macrophage colony-stimulating factor (GM-CSF) (22), and tumor necrosis factor ␣ (TNF␣) (23,24), all of which exhibit elevated protein levels in SLE serum. We also examined effects of anti-IFN␣ mAb on B lymphocyte stimulator/BAFF, which is overexpressed in SLE and is a therapeutic target (25,26); B cell autoantibody production and effector functions are considered so crucial to SLE that BAFF inhibitors are under development to decrease B lymphocyte populations and thereby allow healthy B cells to regenerate to normal levels after treatment. PATIENTS AND METHODS SLE patients and controls. The MI-CP126 trial was a multicenter randomized (2:1), double-blind, placebocontrolled, single-dose, dose-escalation study in patients with SLE, with an open-label extension. The primary objectives were determination of safety and pharmacokinetics. Other trial objectives reported herein included assessment of the effects of anti-IFN␣ mAb on pharmacodynamic markers and assessment of its effects on disease activity. Adults (age ⱖ18 years) who met the American College of Rheumatology criteria for SLE (27,28) were enrolled in the trial. Stable background SLE treatment with acetaminophen, nonsteroidal antiinflammatory drugs, antimalarial agents, and/or prednisone ⱕ20 mg/day (or equivalent) was allowed. Patients receiving cyclophosphamide, azathioprine, methotrexate, mycophenolate mofetil, cyclosporine, prednisone ⬎20 mg/day (or equivalent), immunoglobulin, blood products, investigational drugs, or antiviral therapies were excluded. Also excluded were patients with active or chronic infection, recent vaccination with live attenuated viruses, recent herpes zoster virus infection, history of severe herpesvirus infection, active central nervous system lupus, clinically significant cardiac, cerebrovascular, liver, or renal disease, or history of cancer. Most of the patients were middle-aged white women with mild to moderately active SLE with cutaneous involvement. The MI-CP126 trial was conducted in accordance with the Declaration of Helsinki, and the study protocol was approved by the institutional review board at each site. All patients provided written informed consent before studyrelated procedures were performed. Subjects were treated with anti-IFN␣ mAb in single escalating intravenous doses of 0.3, 1.0, 3.0, 10.0, or 30.0 mg/kg. All patients were followed up for 84 days. A total of 62 patients were enrolled in the trial (3 patients participated in both the blinded and the open-label portions). The ages of patients ranged from 23 to 80 years, and the female:male ratio was ⬃20:1. Whole blood samples for IFN␣-inducible gene expression profiling were collected in PAXgene RNA tubes (PreAnalytiX, Hilden, Germany) on study day 0 (before dosing) and on ANTI-IFN␣ TREATMENT OF SLE postdosing days 1, 2, 4, 7, 14, 28, and 84. Skin biopsy samples were collected on study day 0 and on postdosing day 14. Skin biopsy specimens were preferentially obtained from involved skin, with followup specimens obtained from near the original biopsy site when possible. Control whole blood samples were obtained from 24 healthy donors enrolled internally (MedImmune) (ages 26–56 years; female:male ratio 5:1). The majority of these donors were white. Blood was collected in PAXgene RNA tubes. All healthy donors provided written informed consent. Total RNA extraction, microarray processing, and microarray data analysis. The Human Genome U133 Plus 2.0 array platform (Affymetrix, Santa Clara, CA) was used to evaluate the effects of anti-IFN␣ mAb in whole blood from the 62 SLE patients and in lesional skin from the 16 patients from whom skin biopsy samples were collected. The general procedures for sample processing and data analysis for microarray studies have been described previously (19,20). Other methods. Whole blood from healthy donors was stimulated ex vivo. Cytokine gene signature scores were calculated. Neutralization of IFN␣/␤-inducible genes in whole blood from patients with SLE was measured, and genes affected by anti-IFN␣ mAb were ranked. Details on these procedures, as well as on immunohistochemistry analysis and other experimental methods, are available online at http:// www.medimmune.com/translationalscience/data/MI-CP126A&R-2009. RESULTS Anti-IFN␣ mAb neutralizes overexpression of IFN␣/␤-inducible genes in whole blood from SLE patients in a dose-dependent manner. To evaluate whether anti-IFN␣ mAb affects its target in SLE patients prior to their receiving anti-IFN␣ mAb treatment, we profiled whole blood from 62 SLE patients (3 patients were enrolled in both the blinded and the open-label portions of the trial). Whole blood samples from all patients were collected before dosing and 1, 2, 4, 7, 14, 28, and 84 days postdosing. We first evaluated the expression of IFN␣/␤inducible genes in SLE. Because IFN␣/␤ protein levels were difficult to measure in SLE patients, IFN␣/␤ activity in whole blood was evaluated using IFN␣/␤inducible gene signature scores (Figure 1A). The scores indicated that IFN␣/␤-inducible genes were overexpressed in whole blood from a majority of the SLE patients enrolled in the trial, and that the gene signature score in patients was significantly increased compared with the score in 24 healthy controls (mean and median scores in patients 8.4 and 5.4, respectively; both P ⬍ 0.01 versus controls). The magnitude of overexpression of the IFN␣/␤inducible gene signature in whole blood allowed us to categorize patients with SLE as having high, moderate, or weak overexpression of IFN␣/␤-inducible genes (19). 1787 It is likely that more accurate assessment of drug target neutralization would be obtained in patients with high or moderate overexpression (gene signature score of ⱖ4). On study day 0, whole blood samples from 37 of the 62 patients (60%) exhibited high or moderate overexpression of the IFN␣/␤-inducible gene signature. Using this group of 37 patients, we monitored the pharmacodynamic effect of anti-IFN␣ mAb on target neutralization in SLE. Ten of the 37 patients received placebo, and the remaining 27 received anti-IFN␣ mAb in varying single doses (0.3 mg/kg [n ⫽ 5], 1.0 mg/kg [n ⫽6], 3.0 mg/kg [n ⫽ 6], 10.0 mg/kg [n ⫽ 6], or 30.0 mg/kg [n ⫽4]). Data on the anti-IFN␣ mAb target neutralization values on days 1, 2, 4, 7, 14, 28, and 84 postdosing, calculated using the gene signature scores for 21 IFN␣/ ␤-inducible genes in each SLE patient as previously described (19), are available online at http:// www.medimmune.com/translationalscience/data/MICP126-A%26R-2009/. Overall, the IFN␣/␤-inducible gene signature scores in SLE patients who received placebo yielded target neutralization values that oscillated around baseline values for 84 days following treatment. Patients treated with 0.3 mg/kg anti-IFN␣ mAb exhibited a substantial decrease in IFN␣/␤inducible gene signature scores in the first 2 days postdosing (57% and 46% mean target neutralization on day 1 and day 2 post–anti-IFN␣ mAb treatment, respectively). Although IFN␣/␤-inducible gene signature scores in this group gradually recovered over time, on day 84 postdosing we still observed average neutralization of 25%. Furthermore, with increases in the antiIFN␣ mAb dose from 3 mg/kg to 10 mg/kg to 30 mg/kg, there was a dose-dependent increase in target neutralization, especially in the early days posttreatment (days 1, 2, and 4). These data provide evidence that overexpression of IFN␣/␤-inducible gene signatures in whole blood from SLE patients could serve as a potential pharmacodynamic biomarker for the evaluation of antiIFN␣ mAb therapy in SLE (19). The gene signature scores for IFN␣/␤-inducible genes in whole blood as calculated using the list of 21 IFN␣/␤-inducible genes (static list) were compared with scores calculated using alternative lists, i.e., the 25 most highly overexpressed IFN␣/␤-inducible genes in each patient (dynamic lists, potentially differing between patients). The correlation coefficient between results obtained with the 2 score calculation methods was 0.95, suggesting that either algorithm was sufficient to capture the magnitude of overexpression of the IFN␣/␤inducible genes in whole blood from patients with SLE. Data on both scores in each individual patient, as well as on the genes used in the calculation in an individual 1788 YAO ET AL Figure 1. Five cytokine-inducible gene signature scores in whole blood (WB) from systemic lupus erythematosus (SLE) patients and from normal controls. A, Relative expression of the interferon-␣/␤ (IFN␣/␤)–inducible gene signature in whole blood from 62 SLE patients before anti-IFN␣ monoclonal antibody (mAb) treatment and from 24 normal controls. A high IFN␣/␤-inducible gene signature was defined as a score of ⱖ10, and a moderate signature was defined as a score of ⱖ4 and ⬍10. Thirty-seven patients (60%) had a moderate or high IFN␣/␤-inducible gene signature score. B, Relative expression of the gene signature for 4 cytokine-inducible genes (granulocyte– macrophage colony-stimulating factor [GM-CSF], interleukin-10 [IL-10], IL-1␤, and tumor necrosis factor ␣ [TNF␣]) in whole blood from the SLE patients before anti-IFN␣ mAb treatment and from normal controls. The gene signature score is calculated as the median fold change, using the 15 most highly induced cytokine-inducible genes (as determined using each cytokine in a separate ex vivo stimulation experiment), as measured with the Human Genome U133 Plus 2.0 array platform (Affymetrix, Santa Clara, CA). Horizontal bars show the medians. Signature scores for the IFN␣/␤-inducible and IL-10 genes were significantly different between patients and controls (P ⬍ 0.01). patient, are available online at http://www.medimmune. com/translationalscience/data/MI-CP126-A%26R2009/. In this representative patient, 19 of the 21 “static” genes were included among the 25 most overexpressed highly IFN␣/␤-inducible genes, further demonstrating the similarity between the static list and dynamic list approaches. The mean (and SEM) target neutralization values calculated at each time point posttreatment, for each dose level, were calculated using the static list of 21 IFN␣/␤-inducible genes (available online at http:// www.medimmune.com/translationalscience/data/MICP126-A%26R-2009/). Dose-dependent target neutralization was again demonstrated, as evaluated using a criterion of ⬎50% neutralization at any time point. To provide a statistical summary of the differences in target neutralization between placebo-treated patients and pa- tients treated with anti-IFN␣ mAb at each dose level across time, Hotelling’s T2 test was applied. This multivariate analog to Student’s t-test accounts for the correlation structure between the time points posttreatment. The neutralization values for each dose were compared with those obtained with placebo, separately, using data from days 1–14 postdosing, since the half-life of antiIFN␣ mAb is within the range of 14–20 days. From the pairwise comparisons of target neutralization values between patients treated with anti-IFN␣ mAb at each of the 5 dose levels and patients treated with placebo, the effect of anti-IFN␣ mAb at the 3 mg/kg, 10 mg/kg, and 30 mg/kg doses was found to be significantly different from that of placebo (P ⫽ 0.03, 0.01, and 0.02, respectively); target neutralization values in patients treated with anti-IFN␣ mAb at 0.3 mg/kg or 1 mg/kg were not significantly different from values in ANTI-IFN␣ TREATMENT OF SLE 1789 Figure 2. Dose responses to anti-IFN␣ mAb therapy in whole blood from SLE patients with overexpression of the IFN␣/␤-inducible gene signature as determined by the gene signature score. A, Neutralization of 21 IFN␣/␤-inducible genes from day 0 (pretreatment) to day 84 (posttreatment), averaged for each study cohort (placebo [red], anti-IFN␣ mAb 0.3 mg/kg [blue], anti-IFN␣ mAb 1 mg/kg [green], anti-IFN␣ mAb 3 mg/kg [orange], anti-IFN␣ mAb 10 mg/kg [black], and anti-IFN␣ mAb 30 mg/kg [pink]). The fraction of neutralization on each study day was subtracted from 1 for each patient separately. Values that exceed 1 from this formula represent increased transcript levels of IFN␣/␤-inducible genes in whole blood following treatment (mostly observed in placebo-treated patients). Values are the mean ⫾ SEM. B, Principal components analysis (PCA) plots obtained using a static list of 21 IFN␣/␤-inducible genes. PCA showed that placebo treatment did not cause significant change in the IFN␣/␤-inducible gene signature in whole blood from SLE patients. Blue dots represent the 24 normal subjects; red dots represent SLE patients before treatment (day 0). Left plot shows results obtained on day 1 after placebo treatment (cyan); center plot shows results obtained on day 14 after placebo treatment (black); right plot shows results obtained on day 84 after placebo treatment (yellow). See Figure 1 for other definitions. placebo-treated patients (P ⫽ 0.17 and P ⫽ 0.10, respectively). Specificity of neutralization of the IFN␣/␤inducible gene signature by anti-IFN␣ mAb. The IFN␣/ ␤-inducible gene signature was not significantly altered in whole blood from SLE patients treated with placebo (Figure 2A), and the neutralization observed in the anti-IFN␣ mAb–treated patients was dose dependent. These findings indicate that the target neutralization observed in SLE patients following anti-IFN␣ mAb treatment is likely drug specific. A single-factor analysis of variance was used to evaluate the differences in IFN␣/␤-inducible gene signature scores across all time points. There was not a significant difference in the IFN␣/␤ gene signature score in the placebo-treated patients (n ⫽ 17) between any days pre- or postdosing (P ⫽ 0.94). Each pairwise time point comparison was also assessed using Tukey’s honest significant difference test; again, no pairwise significant differences were observed. Similar results were obtained (P ⫽ 0.31) in analyses including only the placebo-treated patients who had a baseline IFN␣/␤-inducible gene signature score of ⱖ4 (n ⫽ 10). Figure 2B shows principal components analysis (PCA) plots obtained using the 21 IFN␣/␤inducible genes in whole blood from SLE patients on days 1, 14, and 84 following placebo treatment. Placebo 1790 treatment did not result in significant changes in either direction in the IFN␣/␤-inducible gene signatures in whole blood from SLE patients at any time following treatment. A heat map and PCA plot depicting target neutralization in whole blood from a representative SLE patient treated with a single-dose intravenous injection of 30 mg/kg anti-IFN␣ mAb are available online at http://www.medimmune.com/translationalscience/data/ MI-CP126-A%26R-2009/. In this patient, strong neutralization (81%) was observed on day 1 postdosing, and peak neutralization (98%) occurred on day 4; neutralization diminished in subsequent measurements. Substantial target neutralization (58%) was still observed on day 84 postdosing. PCA showed that the patient’s IFN␣/ ␤-inducible gene signature was significantly decreased on day 1 following drug treatment, was decreased further, to a level comparable with that in healthy controls, on day 4, and then rose steadily from day 14 to day 84 postdosing. We ranked all genes that were neutralized by anti-IFN␣ mAb or changed in placebo-treated patients and found that IFN␣/␤-inducible genes comprised 93 of the top 100 probes neutralized by anti-IFN␣ mAb on day 7 postdosing. In contrast, only 1 of the top 100 probes neutralized with placebo treatment on day 7 postdosing was an IFN␣/␤-inducible gene. The difference was significant (P ⬍ 0.01 by 2-sample [2-tailed] proportions test), suggesting that the effect of anti-IFN␣ mAb was drug-specific in the SLE patients. Similar results were observed on days 1, 2, 4, 14, and 28 postdosing. Neutralization of IFN␣/␤-inducible genes at the disease site is confirmed by transcript profiling and immunohistochemistry. To examine whether target neutralization in whole blood correlated with target neutralization at disease sites, we profiled skin lesions from the 16 SLE patients from whom skin biopsy samples were available, using microarrays. Skin lesion specimens were collected on day 0 predosing and day 14 postdosing. Forty-two of the 50 most highly overexpressed genes in lesional skin from SLE patients were IFN␣/␤-inducible genes, consistent with the observation that the overwhelming majority of genes overexpressed in whole blood from patients with SLE are IFN␣/␤-inducible genes. A list of the 50 most highly overexpressed genes in lesional skin from these 16 SLE patients, and the pretreatment IFN␣/␤-inducible gene signature scores in skin lesions and whole blood from each individual patient, are available online at http://www.medimmune. com/translationalscience/data/MI-CP126-A%26R2009/. Overall, expression patterns of IFN␣/␤-inducible genes were similar in whole blood and skin lesions. In 13 YAO ET AL of the 16 patients, IFN␣/␤-inducible gene signature scores in whole blood and skin were either both above the cutoff for defining presence of the signature (i.e., ⱖ4) or both below the cutoff (P ⬍ 0.05 by Fisher’s exact test). These similar trends of overexpression of IFN␣/␤inducible genes in whole blood and skin lesions from patients with SLE provided further scientific rationale for using whole blood as a surrogate tissue to measure the pharmacodynamics of anti-IFN␣ mAb. We also compared target neutralization trends in skin and whole blood from SLE patients. Of the 8 patients who exhibited positive IFN␣/␤-inducible gene signatures in both skin and whole blood, 7 showed a similar trend of target neutralization in both whole blood and skin lesions on day 14 after receiving either anti-IFN␣ mAb treatment or placebo. These results provided evidence that antiIFN␣ mAb is able to neutralize its target in disease tissue. To determine whether the highly overexpressed IFN␣/␤-inducible genes in lesional skin were associated with similar changes in protein expression, we performed immunohistochemistry analyses to assess the presence of 3 IFN␣/␤-inducible proteins, hect domain and RCC1-like domain 5 (HERC-5), interferon-induced protein 15 (ISG-15), and chemokine (CXC) motif ligand 10 (IP-10). These proteins were chosen based on strong overexpression of their respective mRNA at disease sites and the availability of immunohistochemistry reagents. Skin lesions from the same biopsy samples as were used in the whole-genome array analysis were also used for immunohistochemistry. Immunohistochemistry characterization of the cellular infiltrates (plasmacytoid dendritic cells [pDCs], myeloid DCs [mDCs], and CD4⫹ cells) allowed us to compare numbers of IFN-producing cells and inflammatory cells in paired biopsy samples of lesional skin, obtained predosing and on day 14 postdosing. In selected patients whose paired biopsy specimens were evaluated, lesional skin contained increased numbers of CD4⫹ cells and exhibited significant upregulation of HERC-5, IP-10, and ISG-15 proteins in the dermis. In contrast, skin biopsy samples from normal donors did not contain appreciable numbers of pDCs or mDCs, and did not stain for HERC-5, IP-10, or ISG-15 (results not shown). Figure 3 shows amelioration of SLE skin lesions and decreases in CD4⫹ cell infiltrates and cells expressing IFN-inducible proteins on day 14 postdosing, in a patient who was treated with 10 mg/kg anti-IFN␣ mAb. Immunohistochemical analysis of paired biopsy specimens from lesional skin (day 0 predosing and day 14 postdosing) (Figure 3A) demonstrated significant de- ANTI-IFN␣ TREATMENT OF SLE 1791 Figure 3. Effects of anti-IFN␣ mAb treatment in a single SLE patient who showed a response to treatment with 10 mg/kg anti-IFN␣ mAb. A, Immunohistochemical analyses of paired biopsy specimens from lesional skin (obtained predosing [day 0] and on day 14 postdosing). BDCA2 is a specific marker for plasmacytoid dendritic cells (DCs), CD83 is a marker for myeloid DCs, and CD4 is present on T cells and DCs. Overexpression of hect domain and RCC1-like domain 5 (HERC-5), interferon-induced protein 15 (ISG-15), and chemokine (CXC) motif ligand 10 (IP-10) that was observed in lesional skin on day 0 was decreased on day 14 (original magnification ⫻ 300). A principal components analysis (PCA) plot of data from the same SLE patient, obtained using the 21 IFN␣/␤-inducible genes in the skin lesion compared with a normal population (blue), is also shown; data were obtained before dosing (red) and on day 14 (black). B, Resolution of a skin lesion in the patient following anti-IFN␣ mAb treatment. C, Heat map representation of the neutralization of 21 IFN␣/␤-inducible genes in whole blood from the SLE patient following anti-IFN␣ mAb treatment. Columns left-to-right correspond to day 0 and days 1, 7, 14, and 28 following treatment; rows correspond to the 21 static IFN␣/␤-inducible genes in this patient (predosing). See Figure 1 for other definitions. creases in CD83 and CD4 staining after treatment. Staining for ISG-15, HERC-5, and IP-10 proteins was also significantly reduced on day 14 postdosing (Figure 3A), consistent with observed substantial decreases in levels of mRNA for these respective genes. The PCA plot showing overexpression of the 21 IFN␣/␤-inducible genes in the skin lesion from this SLE patient, compared with normal controls, indicated substantial neutralization of the IFN␣/␤-inducible gene signature on day 14 following anti-IFN␣ mAb treatment. As seen in Figure 3B, the skin lesion also resolved after administration of anti-IFN␣ mAb. A heat map representation of the neutralization of the 21–IFN␣/␤-inducible gene signature in whole blood from this anti-IFN␣ mAb–treated SLE patient is shown in Figure 3C. Strong target neutralization was observed in whole blood from day 1 to day 28 postdosing, consistent with the substantial target neutralization observed in the skin lesion. Similar decreases in levels of inflammatory cells and IFN-inducible proteins were not 1792 YAO ET AL Figure 4. Effects of anti-IFN␣ mAb on cytokine signaling pathways in whole blood from SLE patients predosing and on day 14 postdosing. Each row indicates the overexpression of a different cytokine-inducible gene signature; each vertical column of symbols represents an individual patient. Elevated levels of cytokine gene signatures in each patient are represented by colors approaching red, while colors approaching blue represent low cytokine gene signature values. See Figure 1 for definitions. observed in either placebo-treated patients or anti-IFN␣ mAb–treated patients with no substantial target neutralization in the skin lesion (data available online at http://www.medimmune.com/translationalscience/data/ MI-CP126-A%26R-2009/). Effects of anti-IFN␣ mAb on BAFF, other cytokines, and their pathways in patients with SLE. After demonstrating that gene signature scores enabled measurement of the effects of anti-IFN␣ mAb on IFN␣/␤inducible genes, we next evaluated the effects of antiIFN␣ mAb on the signaling pathways of other cytokines of interest in whole blood from SLE patients. In these experiments, cytokine-inducible gene signature scores were measured using panels of cytokine-inducible genes specific to each cytokine studied. The cytokine-inducible gene signatures were used to evaluate the effect of these cytokines, since the proteins were difficult to measure, similar to the case with IFN␣ protein as described above. Prior to anti-IFN␣ mAb treatment, IFN␣/␤inducible gene signature scores in whole blood (Figure 1A) demonstrated high levels of overexpression in most of the SLE patients. Similarly, prior to anti-IFN␣ mAb treatment, cytokine-inducible gene signature scores for GM-CSF, TNF␣, IL-1␤, and IL-10 indicated elevated levels of expression in whole blood in some SLE patients (Figure 1B). Figure 4 shows the effects of anti-IFN␣ mAb and placebo on cytokine signaling pathways in SLE whole blood as measured by cytokine-inducible gene signature scores predosing and on day 14 postdosing. In some patients, suppression of other cytokine-inducible genes was observed along with neutralization of IFN␣/ ␤-inducible genes by anti-IFN␣ mAb. Patient 41, for example, had exceptionally high baseline levels of IFN␣/␤, GM-CSF, TNF␣, and IL-1␤ activity, as demonstrated by cytokine-inducible gene signature scores (Figure 4). The PCA plots in Figure 5 show the effects of anti-IFN␣ mAb on IFN␣/␤-, IFN␥-, TNF␣-, and IL-1␤– inducible genes in whole blood from this patient on days 1, 14, and 84 postdosing. The overexpressed IFN␣/␤inducible gene signature exhibited a rapid decrease on day 1 postdosing, remained low on day 14, and increased substantially on day 84 (Figure 5A). A similar trend in the IFN␥-inducible gene signature in patient 41 was observed following anti-IFN␣ mAb treatment (Figure 5B). Both the TNF␣- and the IL-1␤–inducible gene signatures decreased to levels comparable with those in healthy controls on day 1 postdosing, then recovered significantly on day 14 and were maintained near the day-14 level on day 84 postdosing (Figures 5C and D). Ex vivo IFN␣ stimulation study of healthy donor whole blood indicated that BAFF is inducible by IFN␣. Prior to treatment with anti-IFN␣ mAb, significant overexpression of BAFF mRNA was observed in whole ANTI-IFN␣ TREATMENT OF SLE 1793 Figure 5. Principal components analysis (PCA) plots of data from SLE patient 41 (treated with a single dose of 10 mg/kg anti-IFN␣ mAb) and a normal population (blue), obtained using the 15 (25 for IFN␣/␤) most highly overexpressed cytokine-inducible genes in whole blood from the patient. A, IFN␣/␤. B, IFN␥. C, TNF␣. D, IL-1␤. Data on the SLE patient were obtained before dosing (red) and on days 1 (cyan), 14 (black), and 84 (yellow) postdosing. See Figure 1 for other definitions. blood from some SLE patients as compared with normal controls, as measured by TaqMan quantitative real-time PCR. Data on relative expression levels of mRNA for BAFF, GM-CSF, IL-10, TNF␣, and IL-1␤ in whole blood and lesional skin from SLE patients versus controls are available online at http://www.medimmune. com/translationalscience/data/MI-CP126-A%26R2009/. Anti-IFN␣ mAb treatment suppressed BAFF mRNA expression in blood (data not shown), as well as expression of mRNA for BAFF, GM-CSF, and TNF␣ in lesional skin (Figure 6). Changes in expression levels of IFN␣/␤-inducible genes in SLE skin lesions were posi- 1794 YAO ET AL Figure 6. Relative expression of mRNA for BAFF, GM-CSF, and TNF␣ in lesional skin from SLE patients before dosing (day 0) and 14 days after treatment with placebo (A) or anti-IFN␣ mAb (B). Expression was measured by TaqMan quantitative real-time polymerase chain reaction and compared with that in pooled samples from normal donors. Only patients who had ⬎2-fold overexpression of the transcripts either predosing or postdosing (or both) are included. One patient treated with placebo exhibited a reduction in expression of mRNA for BAFF, GM-CSF, and TNF␣ in lesional skin, along with a reduction in IFN␣/␤-inducible genes. See Figure 1 for definitions. tively correlated with the changes in mRNA for GM-CSF, TNF␣, and BAFF ( ⫽ 0.67, 0.92, and 0.82, respectively). DISCUSSION We are currently exploring the use of an antiIFN␣ monoclonal antibody as therapy for SLE. As part of this effort, we are testing a genomic approach to develop pharmacodynamic biomarkers that would aid in monitoring anti-IFN␣ mAb activity in clinical trials and to inform dosage selection in subsequent trials. In this study, we have shown that neutralization of a gene signature comprising 21 IFN␣/␤-inducible genes in whole blood from SLE patients can serve as a pharmacodynamic biomarker for assessing anti-IFN␣ mAb activity. Our studies demonstrate that this biomarker possesses many characteristics of an ideal biomarker (18). It is sensitive and specific. In SLE whole blood, it is stable and can be quantitatively measured with multiple assays. This static 21-gene signature panel captures the overexpression of the IFN␣/␤-inducible genes in whole blood from SLE patients and allows the categorization of patients as having high, moderate, or weak overexpression of IFN␣/␤-inducible genes. Using these 21 genes, we have demonstrated a specific and dosedependent neutralization of IFN␣ by anti-IFN␣ mAb. This pharmacodynamic marker enables measurement of ANTI-IFN␣ TREATMENT OF SLE biologic activity of anti-IFN␣ mAb in an easily accessible surrogate tissue, whole blood. In addition to allowing sampling at multiple time points in a relatively noninvasive and cost-effective manner, whole blood is the surrogate tissue most frequently used to monitor pharmacokinetics in clinical trials. Thus, pharmacokinetic/ pharmacodynamic modeling using IFN␣/␤-inducible gene signature scores along with other factors (such as clinical benefit) can guide dose scheduling for anti-IFN␣ mAb in future SLE trials. We also used lists of the 25 most overexpressed IFN␣/␤-inducible genes in individual SLE patients to evaluate the magnitude of overexpression of IFN␣ in these patients. This approach should enable more accurate evaluation of the activation of the IFN␣/␤ signaling pathway in patients in whom the most highly overexpressed IFN␣/␤-inducible genes are not included in the 21-gene panel. Use of a static 21-gene panel is statistically more robust than use of a patient-specific dynamic set of 25 genes, and is more feasible than use of a dynamic panel as a potential diagnostic marker to predict response to anti-IFN␣ mAb treatment in individual SLE patients. Notably, in analyses of whole blood from the 62 SLE patients, the correlation between the IFN␣/␤-inducible gene signature scores obtained using 21 genes and those obtained using 25 genes was very strong (r ⫽ 0.95) (scores in individual patients available online at http://www.medimmune.com/translational science/data/MI-CP126-A&R-2009). It is important to note that the IFN␣/␤-inducible gene signatures had not returned to initial baseline levels at the last time point assessed (day 84 postdosing), even with the lowest dose of anti-IFN␣ mAb (0.3 mg/kg). Pharmacokinetic data indicated that drug availability in the serum of the patients was very low on day 84 in the 0.3 mg/kg–treated cohort (half-life of antiIFN␣ mAb was ⬃18 days). The sustained neutralization of IFN␣/␤-inducible genes on day 84 was more significant at a higher drug dose, particularly at the 10 mg/kg and 30 mg/kg doses, where drug was still available in the serum (based on pharmacokinetic data). These changes were observed despite the limited number of SLE patients eligible for pharmacodynamic evaluation (4–6 in each anti-IFN␣ mAb treatment cohort). These studies also showed that anti-IFN␣ mAb neutralization of IFN␣ observed in whole blood as surrogate tissue adequately represented IFN␣ neutralization occurring at disease sites (in this case, skin lesions in patients with mild to moderate SLE). In most cases, microarray analysis data on SLE skin lesions paralleled the data obtained with whole blood. Further, immunohistochemical analysis revealed that treated pa- 1795 tients whose symptoms responded to therapy all showed significant decreases in levels of inflammatory and selected IFN␣/␤-inducible proteins. These changes were not observed in the majority of the placebo-treated patients evaluated by immunohistochemistry. It should be noted, however, that the sample size was too small to draw statistically robust conclusions based on data from the MI-CP126 trial. Also, obtaining quality skin biopsy specimens pre- and post–anti-IFN␣ mAb treatment posed some technical challenges. However, our preliminary data are encouraging, and we will continue to generate more data in ongoing trials in which patients with severe SLE will also be enrolled. While we focused on using IFN␣/␤-inducible genes to evaluate the pharmacodynamic effect of antiIFN␣ mAb in SLE, the power of whole-genome microarray analysis for such studies allowed us also to examine the effect of anti-IFN␣ mAb on other signaling pathways in SLE. In the MI-CP126 trial, we evaluated how anti-IFN␣ mAb treatment affected BAFF and other signaling pathways (GM-CSF, IFN␥, IL-10, TNF␣, and IL-1␤) in the periphery and lesional skin of patients with SLE. These signaling pathways were activated in selected patients, as evidenced by overexpression of their transcripts and/or cytokine-inducible genes in peripheral blood. The transcripts of these signaling molecules and/or their inducible gene signatures showed trends for change similar to those observed with IFN␣/ ␤-inducible genes, which suggests that these pathways may reside downstream of IFN␣ in SLE, at least in some patients. Understanding the broader importance of this observation with regard to SLE pathogenesis and therapy will require further research. It should be noted that the patients in this trial represent only one subpopulation of SLE patients, specifically those with mild to moderate SLE with skin involvement. The presence and usefulness of this 21– IFN␣/␤-inducible gene signature is currently being tested in patients with more severe skin involvement and other manifestations of disease (i.e., lupus nephritis, neurologic abnormalities). More patients are being evaluated in the subsequent trial so that statistically more meaningful conclusions about this pharmacodynamic gene signature biomarker can be drawn. It remains to be determined whether this gene signature could be used to prospectively identify patients who are most likely to benefit from anti-IFN␣ mAb therapy and to help optimize their dosing. Current and future work aims to further evaluate potential correlations between overexpression of the IFN␣/␤-inducible gene signature in whole blood from SLE patients and clinical benefits of 1796 YAO ET AL anti-IFN␣ mAb, to develop models for predicting patient response to treatment. ACKNOWLEDGMENTS We would like to thank Anmarie Boutrin, Nancy Huddy, and Martha Wester for technical assistance, Jiaqi Huang and Robert Georgantas III for critical review of the manuscript, and Eric Phan, Krystal Bowers, and Denise Dawson for clinical trial sample management. 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. Yao 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. Yao, Richman, Morehouse, White, Kiener, Jallal. Acquisition of data. Yao, Richman, Morehouse, de los Reyes, Brohawn. Analysis and interpretation of data. Yao, Higgs, Morehouse, Zhang, Coyle, Kiener, Jallal. REFERENCES 1. American College of Rheumatology Ad Hoc Committee on Systemic Lupus Erythematosus Guidelines. Guidelines for referral and management of systemic lupus erythematosus in adults. Arthritis Rheum 1999;42:1785–96. 2. Vasoo S, Hughes GR. Theory, targets and therapy in systemic lupus erythematosus. Lupus 2005;14:181–8. 3. Smith PL, Lombardi G, Foster GR. Type I interferons and the innate immune response: more than just antiviral cytokines. Mol Immunol 2005;42:869–77. 4. Conlon KC, Urba WJ, Smith JW II, Steis RG, Longo DL, Clark JW. Exacerbation of symptoms of autoimmune disease in patients receiving ␣-interferon therapy. Cancer 1990;65:2237–42. 5. Fattovich G, Giustina G, Favarato S, Ruol A. A survey of adverse events in 11,241 patients with chronic viral hepatitis treated with alfa interferon. J Hepatol 1996;24:38–47. 6. Funk J, Langeland T, Schrumpf E, Hanssen LE. Psoriasis induced by interferon-␣. Br J Dermatol 1991;125:463–5. 7. Okanoue T, Sakamoto S, Itoh Y, Minami M, Yasui K, Sakamoto M, et al. Side effects of high-dose interferon therapy for chronic hepatitis C. J Hepatol 1996;25:283–91. 8. Kirou KA, Lee C, George S, Louca K, Peterson MG, Crow MK. Activation of the interferon-␣ pathway identifies a subgroup of systemic lupus erythematosus patients with distinct serologic features and active disease. Arthritis Rheum 2005;52:1491–503. 9. Baechler EC, Batliwalla FM, Karypis G, Gaffney PM, Ortmann WA, Espe KJ, et al. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc Natl Acad Sci U S A 2003;100:2610–5. 10. Bennett L, Palucka AK, Arce E, Cantrell V, Borvak J, Banchereau J, et al. Interferon and granulopoiesis signatures in systemic lupus erythematosus blood. J Exp Med 2003;197:711–23. 11. Han GM, Chen SL, Shen N, Ye S, Bao CD, Gu YY. Analysis of gene expression profiles in human systemic lupus erythematosus using oligonucleotide microarray. Genes Immun 2003;4:177–86. 12. Hua J, Kirou K, Lee C, Crow MK. Functional assay of type I interferon in systemic lupus erythematosus plasma and association with anti–RNA binding protein autoantibodies. Arthritis Rheum 2006;54:1906–16. 13. Bengtsson AA, Sturfelt G, Truedsson L, Blomberg J, Alm G, Vallin H, et al. Activation of type I interferon system in systemic lupus erythematosus correlates with disease activity but not with antiretroviral antibodies. Lupus 2000;9:664–71. 14. Dall’Era MC, Cardarelli PM, Preston BT, Witte A, Davis JC Jr. Type I interferon correlates with serological and clinical manifestations of SLE. Ann Rheum Dis 2005;64:1692–7. 15. Hooks JJ, Moutsopoulos HM, Geis SA, Stahl NI, Decker JL, Notkins AL. Immune interferon in the circulation of patients with autoimmune disease. N Engl J Med 1979;301:5–8. 16. Pascual V, Farkas L, Banchereau J. Systemic lupus erythematosus: all roads lead to type I interferons. Curr Opin Immunol 2006;18: 676–82. 17. Wallace DJ, Petri M, Olsen N, Kirou K, Dennis G, Yao Y, et al. MEDI-545, an anti-interferon alpha monoclonal antibody, shows evidence of clinical activity in systemic lupus erythematous [abstract]. Arthritis Rheum 2007;56 Suppl 9:S526–7. 18. Adjei AA. Incorporating biomarkers into early phase clinical trials [abstract]. Ann Oncol 2006;17 Suppl 3:iii21. 19. Yao Y, Higgs BW, Morehouse C, de los Reyes M, Trigona W, Brohawn P, et al. Development of potential pharmacodynamic and diagnostic markers for anti-IFN-␣ monoclonal antibody trials in systemic lupus erythematosus. Hum Genomics Proteomics, 2009. URL: http://www.sage-hindawi.com/getarticle.aspx?doi⫽10.4061/ 2009/374312. 20. Yao Y, Richman L, Morehouse C, de los Reyes M, Higgs BW, Boutrin A, et al. Type I interferon: potential therapeutic target for psoriasis? PLoS ONE 2008;3:e2737. 21. Chun HY, Chung JW, Kim HA, Yun JM, Jeon JY, Ye YM, et al. Cytokine IL-6 and IL-10 as biomarkers in systemic lupus erythematosus. J Clin Immunol 2007;27:461–6. 22. Fiehn C, Wermann M, Pezzutto A, Hufner M, Heilig B. Plasma GM-CSF concentrations in rheumatoid arthritis, systemic lupus erythematosus and spondyloarthropathy. Z Rheumatol 1992;51: 121–6. In German. 23. Borodin AG, Baranov AA, Kliukvina NG, Abaitova NE, Nasonov EL. Clinical and pathogenic significance of tumor necrosis factor-␣ in systemic lupus erythematosus. Ter Arkh 2002;74:32–5. In Russian. 24. Palucka AK, Blanck JP, Bennett L, Pascual V, Banchereau J. Cross-regulation of TNF and IFN-␣ in autoimmune diseases. Proc Natl Acad Sci U S A 2005;102:3372–7. 25. Morimoto S, Nakano S, Watanabe T, Tamayama Y, Mitsuo A, Nakiri Y, et al. Expression of B-cell activating factor of the tumour necrosis factor family (BAFF) in T cells in active systemic lupus erythematosus: the role of BAFF in T cell-dependent B cell pathogenic autoantibody production. Rheumatology (Oxford) 2007;46:1083–6. 26. Stohl W, Metyas S, Tan SM, Cheema GS, Oamar B, Xu D, et al. B lymphocyte stimulator overexpression in patients with systemic lupus erythematosus: longitudinal observations. Arthritis Rheum 2003;48:3475–86. 27. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield NF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982;25:1271–7. 28. Hochberg MC, for the Diagnostic and Therapeutic Criteria Committee of the American College of Rheumatology. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus [letter]. Arthritis Rheum 1997;40:1725.