SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To examine the relationship between circulating B lymphocyte stimulator (BLyS) levels and humoral responses to influenza vaccination in systemic lupus erythematosus (SLE) patients, as well as the effect of vaccination on BLyS levels, and to investigate clinical and serologic features of SLE that are associated with elevated BLyS levels.

Methods

Clinical history, disease activity measurements, and blood specimens were collected from 60 SLE patients at baseline and after influenza vaccination. Sera were tested for BLyS levels, lupus-associated autoantibodies, serum interferon-α (IFNα) activity, 25-hydroxyvitamin D (25[OH]D), and humoral responses to influenza vaccination.

Results

Thirty percent of the SLE patients had elevated BLyS levels, with African American patients having higher BLyS levels than white patients (P = 0.006). Baseline BLyS levels in patients were not correlated with humoral responses to influenza vaccination (P = 0.863), and BLyS levels increased postvaccination only in the subset of patients with BLyS levels in the lowest quartile (P = 0.0003). Elevated BLyS levels were associated with increased disease activity, as measured by the SLE Disease Activity Index, physician's global assessment, and Systemic Lupus Activity Measure in white patients (P = 0.035, P = 0.016, and P = 0.018, respectively), but not in African Americans. Elevated BLyS levels were also associated with anti–nuclear RNP (P = 0.0003) and decreased 25(OH)D (P = 0.018). Serum IFNα activity was a significant predictor of elevated BLyS in a multivariate analysis (P = 0.002).

Conclusion

Our findings indicate that African American patients with SLE have higher BLyS levels regardless of disease activity. Humoral response to influenza vaccination is not correlated with baseline BLyS levels in SLE patients, and only those patients with low baseline BLyS levels demonstrate an increased BLyS response after vaccination.

B lymphocyte stimulator (BLyS), also known as tumor necrosis factor (TNF) ligand superfamily member 13B and BAFF, is a type II transmembrane protein and a member of the TNF superfamily (1, 2). BLyS is produced by several different cell types, including monocytes, activated neutrophils, T cells, and dendritic cells (3–5), and is expressed as a cell surface protein which can be furin-cleaved and released into the circulation. Although BLyS has been shown to be constitutively expressed, certain inflammatory cytokines, such as interleukin-2, TNFα, interferon-γ (IFNγ), and IFNα, can enhance its production and secretion (3, 4, 6, 7). BLyS binds to 3 receptors found primarily on B cells (8). Activation of these receptors leads to B cell proliferation, differentiation, survival, and IgG class switching (1, 4, 9). BLyS has been shown to play an important role in primary immune responses, as evidenced by the fact that anti-BLyS–treated mice show profoundly reduced numbers of naive B cells with accompanying attenuated responses to both T cell–dependent and T cell–independent antigens (10).

Transgenic mice that overexpress BLyS display a myriad of autoimmune features, such as high levels of rheumatoid factor, anti-DNA, circulating immune complexes, and immunoglobulin deposition in the kidneys (9). Mice treated with exogenous BLyS develop increased B cell numbers, particularly those directed against chromatin (11), and autoreactive cells encountering transitional B cell checkpoints in the spleen require higher concentrations of BLyS to survive than do nonautoreactive B cells (8, 9, 12). Mouse models have also demonstrated that deletion of either BLyS or its receptor severely impairs B cell development beyond the transitional stage, with a resulting decrease in peripheral B cell populations (9, 13–15).

Elevated circulating BLyS levels have been found in patients with systemic autoimmune disorders such as systemic lupus erythematosus (SLE), rheumatoid arthritis, and Sjögren's syndrome (16–18). Early studies found increased serum BLyS levels in SLE patients compared to healthy individuals that correlated with anti–double-stranded DNA (anti-dsDNA) titers, but not disease activity (16, 17, 19). BLyS levels did, however, decrease following high-dose corticosteroid treatment (19). Other studies found that peripheral blood leukocyte BLyS messenger RNA (mRNA) levels correlated with SLE disease activity (20, 21).

While most of these initial studies, which were performed in mainly Hispanic patients with SLE, failed to demonstrate a correlation between serum BLyS levels and disease activity, a later longitudinal study that was undertaken at 4 different clinical centers and contained ancestral diversity found an association between increases in plasma BLyS levels from a previous visit and increases in scores on the Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA) version of the SLE Disease Activity Index (SLEDAI) at the next visit (22). Additionally, a study involving Norwegian patients with SLE found a correlation between serum BLyS levels and SLEDAI scores (23). A study examining plasma BLyS protein levels in pediatric SLE also demonstrated an association between elevated BLyS levels and increased disease activity (24).

IFNα has proven to be a key cytokine in SLE pathogenesis and has been shown to increase BLyS expression in antigen-presenting cells (4). Therefore, it is possible that IFNα plays a role in driving BLyS production in SLE, although a direct correlation between circulating BLyS levels and serum IFNα activity has yet to be demonstrated in SLE patients. Additionally, several environmental factors have been implicated in the pathogenesis of SLE (25), including vitamin D deficiency (26). Since vitamin D has been shown to suppress the expression of the IFN signature in myeloid-derived dendritic cells, as well as suppress B cell activation and immunoglobulin production (27, 28), it is of interest to assess the relationship between vitamin D and BLyS levels.

Belimumab is a fully human monoclonal antibody that binds BLyS and inhibits its activity. Two recent large, multicenter, randomized, double-blind, placebo-controlled phase III trials, the Study of Belimumab in Subjects with SLE 52-week trial (BLISS-52) and the BLISS 76-week trial (BLISS-76), evaluated the efficacy of belimumab plus standard therapy in comparison to placebo plus standard therapy; positive results were obtained, with both studies reaching their primary efficacy end points of reductions in disease activity without ancillary organ flares (29, 30).

Taken together, these data suggest that excessive BLyS levels may be an important mechanism underlying SLE pathogenesis and that BLyS-targeted therapies seem promising. However, normal B cell function requires some BLyS activity. To date, no studies have evaluated whether specific levels of BLyS are required to mount appropriate vaccination responses or whether BLyS levels change after vaccination. The primary objective of this study is to investigate the relationship between circulating BLyS levels and humoral responses to influenza vaccination, as well as the effect of influenza vaccination on BLyS production in SLE. The present study also examines select demographic, environmental, and clinical features of SLE for association with elevated BLyS levels in an effort to better understand the mechanisms of elevated BLyS in a heterogeneous SLE population.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Study population.

Experiments were performed in accordance with the Helsinki Declaration and were approved by the Institutional Review Boards at the Oklahoma Medical Research Foundation (OMRF) and the University of Oklahoma Health Sciences Center. All of the study participants provided informed consent prior to enrollment. Sixty female patients who met at least 4 of the American College of Rheumatology (ACR) classification criteria for SLE (31, 32) and were enrolled in a longitudinal influenza study cohort at the OMRF were included in this study. Additionally, 60 healthy individuals were recruited to provide control samples and were matched to the patients based on age (±5 years), race, and sex.

Data collection.

Participants provided demographic information that included sex, age, and self-reported race. Clinical information was extracted from the medical records using the Lupus Family Registry and Repository collection tool for ACR classification criteria (33), age at diagnosis, and medication use. Peripheral blood was obtained from each participant before vaccination, and 2, 6, and 12 weeks after vaccination. Postvaccination time points were selected for determination of a variety of influenza vaccination immune responses, including predicted maximal T cell reactivity (2 weeks), humoral responses (6 weeks), as well as a more distant time point for potential autoantibody changes (12 weeks). Serum and plasma were isolated and stored at −20°C until used. SLE patients were also evaluated for disease activity by a board-certified rheumatologist at the initial visit and 6 and 12 weeks after vaccination, using the SELENA–SLEDAI, physician's global assessment, and Systemic Lupus Activity Measure (SLAM) (34, 35). The SELENA–SLEDAI flare composite score was computed for the 6- and 12-week visits (36, 37).

Determination of plasma BLyS levels.

Plasma BLyS levels were determined by an enzyme-linked immunosorbent assay (ELISA), according to the recommendations of the manufacturer (R&D Systems). Samples were analyzed in duplicate, and the mean coefficient of variation was 5.3%. Elevated plasma BLyS levels were defined as being higher than the 95th percentile of BLyS levels measured in healthy controls (>1.285 ng/ml).

Serologic measurements.

Plasma 25-hydroxyvitamin D (25[OH]D) levels were determined in duplicate using a commercial enzyme immunoassay according to the recommendations of the manufacturer (Immunodiagnostic Systems). SLE patient blood specimens were tested for complete blood cell count, erythrocyte sedimentation rate (ESR), and creatinine level.

Lupus-associated autoantibodies.

Antinuclear antibodies (ANAs) were detected using a HEp-2 indirect immunofluorescence assay (Inova Diagnostics). Detection of ANAs at a dilution of 1:120 or greater was considered a positive result. Double-stranded DNA antibodies were detected using a Crithidia luciliae indirect immunofluorescence assay (Inova Diagnostics). Detection of anti-dsDNA at a dilution of 1:30 or greater was considered a positive result. ELISAs were used to evaluate sera for antibodies to Sm, nuclear RNP (nRNP), Ro, La, ribosomal P, and cardiolipin as previously described (38, 39). Samples were run in duplicate and normalized to a known positive control. Western blot analysis using HeLa cell extracts was also performed to determine autoantibody specificities (40). Briefly, HeLa cell extracts were subjected to electrophoresis using 12.5% polyacrylamide gels under denaturing conditions and transferred to nitrocellulose. Autoantigens were determined by identifying distinct bands that corresponded to the appropriate molecular weight and to the band from the known positive control serum used in the assay.

Serum IFNα activity.

The reporter cell assay for serum IFNα activity has previously been described in detail (41). Briefly, reporter cells were used to measure the ability of patient sera to up-regulate IFN-induced gene expression. The reporter cells (WISH cells) (catalog no. CCL-25; ATCC) were cultured with 50% patient sera for 6 hours and then lysed. Messenger RNA was purified from cell lysates, and complementary DNA was made from total cellular mRNA and then quantified using real-time polymerase chain reaction. Forward and reverse primers for the genes MX1, PKR, and IFIT, which are known to be highly and specifically induced by IFNα, were used in the reaction (42). Background gene expression was controlled by amplifying GAPDH in the same samples.

Trivalent influenza vaccination.

This study extended over 4 influenza vaccination seasons and included individuals vaccinated with the 2005–2006, 2006–2007, 2007–2008, and 2008–2009 vaccines. All patients received the currently licensed influenza vaccine approved for use in the US.

Anti-influenza vaccine humoral response.

Three measures of vaccine responsiveness, Bmax (relative amounts of native/denatured anti-influenza antibodies), Ka (antibody affinity, the inverse of the dissociation constant Kd), and hemagglutination inhibition, were obtained as previously described (43). A sandwich ELISA was used to quantify antibodies to native glycoproteins, and a nonlinear regression model was used to calculate Bmax and Ka (43). Hemagglutinin assays were performed at 4°C using human red blood cells (43). The sum of the ranks of the 3 measures was used to determine an individual's combined antibody score within each vaccination year.

Statistical analysis.

Categorical variables were analyzed using Fisher's exact test. Normally distributed continuous variables were analyzed using an unpaired t-test, and Welch's correction was used in instances of unequal variance. The Mann-Whitney test was used in instances of non-normality. The false discovery rate (FDR) was used to correct for multiple comparisons. Spearman's correlation was used to assess the relationship between serum IFNα activity, combined antibody scores for vaccination responses, and BLyS levels. Wilcoxon's matched pairs signed rank test was performed to assess changes in BLyS levels after vaccination. Multivariate logistic regression was performed after candidate predictors were identified, and a model with all candidates was fitted and reduced using likelihood ratio tests with data transformation performed as necessary. Potential two-way interactions and confounders were assessed in the multivariate model. Analyses were performed using GraphPad Prism version 5.04 for Windows, SAS version 9.2 (SAS Institute), and NCSS software.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Higher plasma BLyS levels in African American patients with SLE.

Plasma BLyS levels were determined at baseline in all SLE patients and controls. BLyS levels in both groups were non-normally distributed. SLE patients had a higher median BLyS level (1.00 ng/ml [interquartile range (IQR) 0.65–1.51]) than did healthy controls (0.73 ng/ml [IQR 0.64–0.84]) (P = 0.0003 by Mann-Whitney test) (Figure 1A). African American patients with SLE (n = 26) had a higher median plasma BLyS level (1.22 ng/ml [IQR 0.82–2.70]) than white patients with SLE (n = 33) (0.85 ng/ml [IQR 0.57–1.21]) (P = 0.006 by Mann-Whitney test) (Figure 1B). (One of the 60 patients was Hispanic.) No difference in the median plasma BLyS level between African American controls (0.77 ng/ml [IQR 0.61–0.89]) and white controls (0.72 ng/ml [IQR 0.69–0.79]) was noted (P = 0.440 by Mann-Whitney test).

thumbnail image

Figure 1. Increased plasma B lymphocyte stimulator (BLyS) levels at baseline in African American patients with systemic lupus erythematosus (SLE). A, BLyS levels in SLE patients and matched controls. Each symbol represents an individual patient. Bars show the median and interquartile range (IQR). B, BLyS levels in African American (AA) and white patients with SLE. Data are shown as box plots. Each box represents the upper and lower IQR. Lines inside the boxes represent the median. Lines outside the boxes represent the range. P values were determined by Mann-Whitney test.

Download figure to PowerPoint

Elevated BLyS levels were defined as those determined to be higher than the 95th percentile of plasma BLyS levels in healthy controls (>1.285 ng/ml). Eighteen SLE patients (30%) and 3 controls (5%) fell above this cutoff and were classified as having “elevated” BLyS levels. No difference in age between the SLE patients with elevated BLyS levels (mean 40.5 years) and those with normal BLyS levels (mean 43.7 years) was noted (P = 0.345 by Fisher's exact test) (Table 1). Twelve of the patients with elevated BLyS levels were African American (67%), while the remaining 6 were white (33%). The odds ratio (OR) was 4.0 (95% confidence interval [95% CI] 1.2–12.9) (P = 0.024 by Fisher's exact test, FDR q = 0.144) (Table 1).

Table 1. Univariate analysis of demographic characteristics, ACR criteria, disease activity, and medication use in SLE patients with normal BLyS levels and those with elevated BLyS levels*
 Normal BLyS levels (n = 42)Elevated BLyS levels (n = 18)Pq§OR (95% CI)
  • *

    Except where indicated otherwise, values are the percent of patients. OR = odds ratio; 95% CI = 95% confidence interval; CNS = central nervous system; SLEDAI = SLE Disease Activity Index; IQR = interquartile range; SLAM = Systemic Lupus Activity Measure; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index.

  • Defined as B lymphocyte stimulator (BLyS) levels higher than the 95th percentile of BLyS levels in healthy controls.

  • Differences in age were assessed using an unpaired t-test. Categorical variables were assessed using Fisher's exact test. Differences in disease activity scores and prednisone dosages were assessed using the Mann-Whitney test.

  • §

    False discovery rate q value.

  • One patient with a normal BLyS level was Hispanic.

Race     
 White64330.0460.1760.3 (0.1–0.9)
 African American33670.0240.1444.0 (1.2–12.9)
Age, mean years43.740.50.345
Cumulative ACR criteria     
 Malar rash50720.1572.6 (0.8–8.6)
 Discoid rash10440.0040.0567.6 (1.9–30.5)
 Photosensitivity67610.7710.8 (0.3–2.5)
 Oral ulcers64611.0000.9 (0.3–2.7)
 Arthritis88720.1490.4 (0.1–1.4)
 Serositis48780.0460.1763.9 (1.1–13.7)
 Renal disease19610.0020.0426.7 (2.0–22.7)
 CNS disease1200.3100.2 (0.0–3.5)
 Hemolytic anemia5110.5762.5 (0.3–19.3)
 Leukopenia26440.2272.3 (0.7–7.2)
 Lymphopenia24560.0340.1594.0 (1.2–12.9)
 Thrombocytopenia12280.1492.8 (0.7–11.5)
 Immunologic45670.1642.4 (0.8–7.7)
Disease activity measures     
 SLEDAI, median (IQR)2 (0–6)6 (2–9)0.085
 SLAM, median (IQR)8 (5–10)11 (8–14)0.0060.063
 Physician's global assessment, median (IQR)23 (9–49)49 (36–64)0.0210.144
 SDI, median (IQR)1 (1–3)2 (1–4)0.158
Flare (over 12 weeks)38500.4091.6 (0.5–5.0)
Medication     
 Prednisone dosage, median (IQR) mg/day0 (0–5)5 (0–10)0.064
 Azathioprine17220.7191.4 (0.4–5.7)
 Hydroxychloroquine74721.0000.9 (0.3–3.2)
 Mycophenolate mofetil21390.2072.3 (0.7–7.8)
 Methotrexate14111.000.8 (0.1–4.1)
 Cyclophosphamide201.000.9 (0.0–24.2)

Lack of effect of baseline plasma BLyS levels on humoral immune response to influenza vaccination in SLE patients.

Baseline BLyS levels were not correlated with anti-influenza humoral responses as measured by a combined antibody score (Bmax, Ka, and hemagglutination inhibition) following influenza vaccination in SLE patients (Spearman's correlation coefficient r2 = 0.001, P = 0.863) (Figure 2A). There was also no significant correlation between baseline BLyS levels and anti-influenza humoral responses in controls (Spearman's correlation coefficient r2 = 0.045, P = 0.105). Also of interest was the effect of vaccination on circulating BLyS levels. In addition to plasma BLyS levels that were obtained at baseline before vaccination, BLyS levels were determined from samples that were obtained 2 weeks after vaccination. There was no significant change in plasma BLyS levels in SLE patients between baseline and 2 weeks after influenza vaccination (Figure 2B). SLE patients had a median BLyS level of 1.0 ng/ml (IQR 0.6–1.5) at baseline and a median BLyS level of 1.0 ng/ml (IQR 0.7–1.4) 2 weeks after vaccination (P = 0.337 by Wilcoxon's matched pairs signed rank test). However, in the subset of patients with BLyS levels in the lowest quartile at baseline, there was a significant increase in BLyS levels 2 weeks after vaccination (P = 0.0003 by paired t-test) (Figure 2C). This subset of patients had a mean ± SD BLyS level of 0.53 ± 0.07 ng/ml before vaccination and 0.86 ± 0.29 ng/ml after vaccination.

thumbnail image

Figure 2. Lack of association between baseline BLyS levels and influenza vaccination responses in SLE patients. A, Lack of correlation of baseline BLyS levels with influenza vaccination response in SLE patients as measured by a combined antibody score (r2 = 0.001, P = 0.863). The combined antibody score consisted of Bmax, Ka, and hemagglutination inhibition, as described in Patients and Methods. B, BLyS levels at baseline and 2 weeks after vaccination in all 60 SLE patients (P = 0.337). C, BLyS levels at baseline and 2 weeks after vaccination in the subset of SLE patients with baseline BLyS levels in the lowest quartile (n = 15). Each symbol represents an individual patient. See Figure 1 for definitions.

Download figure to PowerPoint

Association of elevated BLyS levels with increased disease activity in white patients but not in African American patients.

White patients with SLE with elevated BLyS levels had higher SLEDAI, physician's global assessment, and SLAM scores than did those with normal BLyS levels (Figures 3A and B). White patients with elevated BLyS levels had a median SLEDAI score of 8 (IQR 5–12), physician's global assessment score of 60 (IQR 39–75), and SLAM of 11 (IQR 9–15), while white patients with normal BLyS levels had a median SLEDAI score of 2 (IQR 0–6), physician's global assessment score of 23 (IQR 9–39), and SLAM of 7 (IQR 5–10) (P = 0.035, P = 0.016, and P = 0.018, respectively, by Mann-Whitney test). African American patients with elevated BLyS levels, however, did not have increased disease activity scores compared to African American patients with normal levels (Figures 3C and D). African American patients with elevated BLyS levels had a median SLEDAI score of 4 (IQR 2–8), physician's global assessment score of 47 (IQR 19–53), and SLAM of 11 (IQR 6–15), while African American patients with normal BLyS levels had a median SLEDAI score of 5 (IQR 2–8), physician's global assessment score of 43 (IQR 12–59), and SLAM of 9 (IQR 6–11) (P = 1.000, P = 0.837, and P = 0.225, respectively, by Mann-Whitney test).

thumbnail image

Figure 3. Association of elevated BLyS levels with increased disease activity in white patients with SLE but not in African American patients with SLE. A and B, Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA) version of the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) (A) and physician's global assessment (PGA) (B) scores in white patients with normal BLyS levels and elevated BLyS levels. C and D, SELENA–SLEDAI (C) and physician's global assessment (D) scores in African American patients with normal BLyS levels and elevated BLyS levels (P = 1.000 for SELENA–SLEDAI score and P = 0.837 for physician's global assessment score in African American patients with normal BLyS levels versus African American patients with elevated BLyS levels). Elevated BLyS levels were defined as being higher than the 95th percentile of levels in healthy matched controls. Symbols represent individual patients. Bars show the median and IQR. See Figure 1 for other definitions.

Download figure to PowerPoint

Association of elevated BLyS levels with discoid rash, renal disease, serositis, and lymphopenia but not with subsequent disease flare.

Patient history with regard to the ACR classification criteria for SLE was compared between SLE patients with elevated BLyS levels and SLE patients with normal BLyS levels (31, 32). In the overall study population, patients had a history of malar rash (57%), discoid rash (20%), photosensitivity (65%), oral ulcers (63%), arthritis (83%), serositis (57%), renal disease (32%), central nervous system disease (8%), hemolytic anemia (5%), leukopenia (28%), lymphopenia (33%), thrombocytopenia (13%), and immunologic features (62%). A medical records review showed that SLE patients with elevated BLyS levels were more likely to meet the following ACR criteria: discoid rash (OR 7.6 [95% CI 1.9–30.5]) (P = 0.004, FDR q = 0.056), renal disease (OR 6.7 [95% CI 2.0–22.7]) (P = 0.002, FDR q = 0.042), serositis (OR 3.9 [95% CI 1.1–13.7]) (P = 0.046, FDR q = 0.176), and lymphopenia (OR 4.0 [95% CI 1.2–12.9]) (P = 0.034, FDR q = 0.159) (by Fisher's exact test) (Table 1). There were no differences between patients with elevated BLyS levels and patients with normal BLyS levels with regard to active disease features that were determined as part of the SLEDAI evaluation, although the number of individuals with each active disease feature was low.

The SELENA–SLEDAI flare composite score was calculated 6 and 12 weeks after baseline plasma BLyS levels were measured (36, 37) (Table 1). Twenty-two individuals experienced a mild/moderate flare at either 6 or 12 weeks after baseline, and 3 individuals experienced a severe flare. Fifty percent of SLE patients with elevated BLyS levels experienced a flare during the 12-week followup period, while 38% of SLE patients with normal BLyS levels at baseline experienced a flare during this time (OR 1.6 [95% CI 0.5–5.0]) (P = 0.409 by Fisher's exact test). The median BLyS level in SLE patients at the visit prior to a flare was 1.14 ng/ml (IQR 0.76–1.50), while the median BLyS level in SLE patients at the time of the flare was 1.00 ng/ml (IQR 0.79–1.55) (P = 0.687 by Wilcoxon's matched pairs signed rank test).

Medication use was also assessed in both groups of SLE patients (Table 1). The median prednisone dosage in patients with normal BLyS levels was 0 mg/day (IQR 0–5), compared to a median of 5 mg/day (IQR 0–10) in patients with elevated BLyS levels (P = 0.064 by Mann-Whitney test). There were no significant differences in the use of azathioprine, hydroxychloroquine, mycophenolate mofetil, or methotrexate between those with elevated BLyS levels and those with normal levels. Only 1 patient was taking cyclophosphamide. Additionally, 1 patient with elevated BLyS levels had received rituximab ∼10 months before the baseline study date, and another patient with normal BLyS levels was receiving infliximab treatment at baseline.

Association of elevated BLyS levels in SLE with anti-nRNP.

SLE patients with elevated BLyS levels had a higher median ANA titer (1:1,080 [IQR 1:360–1:3,240]) than did patients with normal BLyS levels (1:120 [IQR 1:40–1:1,080]) (P = 0.006 by Mann-Whitney test). Patients with elevated BLyS levels also had higher anti-dsDNA titers than patients with normal BLyS levels, although the difference did not reach statistical significance (median titer 1:10 in patients with elevated BLyS levels versus 1:0 in patients with normal BLyS levels; P = 0.061 by Mann-Whitney test). However, there was a significant correlation between anti-dsDNA titers and BLyS levels in SLE patients (Spearman's r2 = 0.15, P = 0.002).

Additional lupus-associated autoantibody specificities were tested in all SLE patients. Twenty-five patients were positive for anti-Ro (42%), 22 patients for anti-nRNP (37%), 20 patients for anti-dsDNA (33%), 12 patients for anti-La (20%), 11 patients for anti-Sm (18%), 9 patients for anticardiolipin (15%), and 8 patients for anti–ribosomal P (13%). When the presence of each autoantibody specificity was assessed categorically, anti-nRNP was significantly associated with elevated BLyS levels, present in 72% of these patients versus only 21% of the patients with normal BLyS levels (OR 9.5 [95% CI 2.7–33.9]) (P = 0.0003 by Fisher's exact test, FDR q = 0.013) (Table 2). The total number of lupus-associated autoantibody specificities was determined for each patient. Those with elevated BLyS levels had a median of 2 autoantibodies (range 0–5), while those with normal BLyS levels had a median of 1 autoantibody (range 0–4) (P = 0.137 by Mann-Whitney test).

Table 2. Univariate analysis of serologic features in SLE patients with normal BLyS levels and those with elevated BLyS levels*
 Normal BLyS levels (n = 42)Elevated BLyS levels (n = 18)Pq§OR (95% CI)
  • *

    Except where indicated otherwise, values are the mean. Anti-dsDNA = anti–double-stranded DNA; anti-nRNP = anti–nuclear RNP; 25(OH)D = 25-hydroxyvitamin D; ESR = erythrocyte sedimentation rate (see Table 1 for other definitions).

  • Defined as BLyS levels higher than the 95th percentile of BLyS levels in healthy controls.

  • Categorical variables were analyzed by Fisher's exact test. Continuous variables were analyzed by unpaired t-test. Welch's correction was used in instances of unequal variance. An ln transformation was performed to correct for non-normality; the Mann-Whitney test was used when ln transformation failed to correct the non-normality.

  • §

    False discovery rate q value.

Autoantibodies, % of patients     
 Anti-dsDNA29440.2492.0 (0.6–6.3)
 Anti-Ro45330.5690.6 (0.2–1.9)
 Anti-La21171.0000.7 (0.2–3.1)
 Anti-nRNP21720.00030.0139.5 (2.7–33.9)
 Anti-Sm14280.2792.3 (0.6–8.9)
 Anti–ribosomal P14111.0000.8 (0.1–4.1)
 Anticardiolipin12220.4312.1 (0.5–9.0)
Serologic features     
 25(OH)D, ng/ml20.616.30.0180.144
 ESR, mm/hour21.135.70.0340.159
 Hematocrit, %37.237.30.531
 Hemoglobin, gm/dl13.217.90.616
 Leukocyte count, ×106 cells/ml6.67.00.933
 Lymphocyte count, ×106 cells/ml2.872.280.058
 Platelet count, ×106 cells/ml2632830.219
 Creatinine, mg/dl0.90.90.390

Western blot analysis using HeLa cell extracts was used to identify the specificities of the nRNP autoantibody production in patients with elevated and normal BLyS levels. Interestingly, anti-nRNP–positive SLE patients with elevated BLyS levels more frequently had autoantibody responses directed against nRNP 70K (67%) than those patients with normal BLyS levels (25%), although this difference was not statistically significant (P = 0.170 by Fisher's exact test). There were no significant associations between the presence of autoantibodies against dsDNA, Ro, La, Sm, ribosomal P, or phospholipids and elevated BLyS levels.

Positive correlation of serum IFNα activity with BLyS levels in SLE.

Serum IFNα activity was measured by a reporter cell assay and was strongly correlated with BLyS levels in SLE patients (Spearman's correlation coefficient r2 = 0.40, P < 0.0001) (Figure 4A). Additionally, SLE patients with elevated BLyS levels (n = 18) had significantly higher median serum IFNα activity (6.7 [IQR 0.5–12.4]) than did patients with normal levels (n = 42) (0.0 [IQR 0.0–0.9]) (P < 0.0001 by Mann-Whitney test) (Figure 4B). African American patients (n = 26) had higher median serum IFNα activity (1.0 [IQR 0.1–9.2]) than did white patients (n = 33) (0.0 [IQR 0.0–1.2]) (P = 0.001 by Mann-Whitney test).

thumbnail image

Figure 4. Positive correlation of serum interferon-α (IFNα) activity with circulating BLyS levels in SLE patients. A, Correlation between IFNα activity and BLyS levels. Circles represent individual African American patients; squares represent individual white patients. Spearman's correlation coefficient was determined. B, Serum IFNα activity in patients with normal and those with elevated BLyS levels. Elevated BLyS levels were defined as being higher than the 95th percentile of levels in healthy matched controls. Bars show the median and IQR. See Figure 1 for other definitions.

Download figure to PowerPoint

Association of elevated BLyS levels with lower 25(OH)D levels and increased ESR in SLE.

Laboratory tests included 25(OH)D level, ESR, complete blood cell count, and serum creatinine level (Table 2). SLE patients with elevated BLyS levels had a lower mean ± SD 25(OH)D level than did patients with normal BLyS levels (16.3 ± 4.8 ng/ml in patients with elevated BLyS levels versus 20.6 ± 8.7 ng/ml in patients with normal BLyS levels) (P = 0.018 by unpaired t-test, FDR q = 0.144). Seventy-two percent of the patients with elevated BLyS levels were considered to be vitamin D deficient (25[OH]D < 20 ng/ml), and 57% of the patients with normal BLyS levels were considered deficient (P = 0.387 by Fisher's exact test). There was no significant difference in the mean ± SD 25(OH)D level between white patients and African American patients (19.2 ± 7.6 ng/ ml in white patients versus 19.3 ± 8.8 ng/ml in African American patients) (P = 0.967 by unpaired t-test).

Patients with elevated BLyS levels also had a higher mean ± SD ESR than patients with normal BLyS levels (35.7 ± 22.1 mm/hour in patients with elevated BLyS levels versus 21.1 ± 13.1 mm/hour in patients with normal BLyS levels) (P = 0.034 by unpaired t-test with Welch's correction, FDR q = 0.159). No significant differences were found between the 2 groups with regard to percent hematocrit, hemoglobin levels, lymphocyte counts, or platelet counts. Both groups of patients had the same mean serum creatinine level (0.9 mg/dl).

Association of increased serum IFNα activity with elevated BLyS levels in a multivariate analysis.

A multivariate logistic regression analysis was performed to determine the relationship between African American ancestry, nRNP autoantibodies, disease activity, serum IFNα activity, 25(OH)D level, and BLyS level. A model with all candidates was fitted and reduced using likelihood ratio tests. The only significant variable in the final model was serum IFNα activity (OR 1.71 [95% CI 1.22–2.39]) (P = 0.002). Two-way interactions between all variables were assessed, and none were found to be significant. African American ancestry and 25(OH)D levels were considered to be potential confounders. Inclusion of either variable did not change the point estimate by greater than 10%, and therefore neither ancestry nor 25(OH)D levels were considered to be confounding.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Several studies have documented elevated circulating levels of BLyS in SLE patients (16, 17, 19, 22) but have suggested inconsistencies in the relationship between increased BLyS levels and disease activity. This is possibly accounted for by differences in the populations studied or the timing or type of disease activity measured. The results of the belimumab trials underscore the likely significance of the BLyS pathway in SLE; however, they also confirm that many patients remain unresponsive to BLyS pathway intervention. The present study explored subsets of SLE patients who may be more likely to have BLyS elevation underlying their disease and investigated whether circulating BLyS levels affected the humoral response to vaccination.

Baseline BLyS levels at vaccination did not correlate with influenza-specific antibody responses in SLE patients. This is an encouraging finding, since BLyS inhibition may be a promising new treatment for SLE, and BLyS has previously been shown to play a role in primary immune responses (10). However, subnormal BLyS levels might be induced in some patients receiving agents antagonistic to the BLyS pathway, and in such cases, poor response to this or other immunizations cannot be excluded. Additionally, vaccination only induced an increase in BLyS production in those patients with low levels at baseline. This phenomenon suggests that patients with high levels at baseline are already producing BLyS at a maximal level and may be restricted from producing additional BLyS by the number of cellular sources or other limiting factors. It is also possible that a certain threshold of BLyS production exists that is sufficient for the formation of primary humoral immune responses and increases in BLyS production occur following vaccination only in those individuals with levels that fall below this threshold.

We found elevated BLyS levels in 30% of the SLE patients in the present study, which is a similar frequency to that reported in other studies looking at a single time point (16, 17). In our study population, we found significantly higher BLyS levels in African American patients with SLE. However, in multivariate analyses, African American race was not selected in the final model. This suggests that factors that are more common in African American patients with SLE, such as nRNP autoantibodies or high IFNα activity (44, 45), may be important BLyS-related variables. Interestingly, the association between elevated BLyS levels and increased disease activity was seen only in the white patients, not in the African American patients, and suggests that the BLyS-mediated pathogenesis and response to BLyS-directed therapies may differ in patients depending on their ancestral background. The presence of elevated IFNα activity and BLyS levels in African American patients may contribute to the poor response to belimumab that was seen in this ancestral subset of patients in phase III clinical trials and are factors that should be considered in the upcoming clinical trial that will focus on the efficacy of the drug in African American patients with SLE (46).

Most previous studies have shown consistent correlations between BLyS protein and mRNA levels and anti-dsDNA titers (16, 17, 19, 22). A significant correlation between BLyS levels and anti-dsDNA titers was also observed in this study, although categorical analysis did not find an increased likelihood of anti-dsDNA positivity in patients with elevated BLyS levels. Additionally, patients with anti-nRNP responses were >9 times as likely to have elevated BLyS levels. These findings highlight the need for additional mechanistic studies examining the impact of BLyS on B cell subsets, including whether there is preferential selection of autoreactive cells that are located in the naive compartment, antigen-exposed B cells present in germinal centers, or both.

In the present study, a significant correlation between BLyS levels and serum IFNα activity was found, and serum IFNα activity was the only significant variable in the multivariate analysis. A similar association between BLyS mRNA expression and a global IFN score in SLE patients has been reported previously (47). Another study examining patients with Sjögren's syndrome after treatment with etanercept demonstrated that serum BLyS levels varied with IFNα activity following therapy (48). Additionally, serum IFNα activity has been shown to strongly correlate with autoantibody production in SLE, and specifically with antibodies against dsDNA and nRNP (45). In vitro, the addition of IFNα to dendritic cells and monocytes results in increased BLyS expression (4). It is therefore reasonable to hypothesize that increased serum IFNα in SLE patients helps to drive BLyS production and that an overlapping spectrum of patients might respond to treatments inhibiting these 2 pathways. Also of interest was the association between elevated BLyS levels and decreased plasma 25(OH)D levels that was independent of racial differences in vitamin D levels. Since vitamin D deficiency is present in approximately two-thirds of SLE patients (49), and has been shown to modulate autoantibody production and the IFN signature (27, 28), it is plausible that vitamin D is capable of either directly or indirectly regulating BLyS levels in SLE.

In conclusion, African American patients with SLE had increased BLyS levels regardless of their disease activity, and BLyS levels were strongly associated with serum IFNα activity in SLE patients. Elevated BLyS levels were also associated with anti-nRNP and decreased 25(OH)D levels. Baseline BLyS levels were not correlated with humoral responses to influenza vaccination in these patients, and only those patients with low BLyS levels at baseline demonstrated an increase in BLyS production after vaccination. These findings support the idea that BLyS is important in SLE disease pathogenesis, although increased levels of BLyS are not necessarily associated with improved humoral responses to vaccination.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

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. James 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. Ritterhouse, Crowe, Thompson, James.

Acquisition of data. Ritterhouse, Niewold, Roberts, Dedeke, James.

Analysis and interpretation of data. Ritterhouse, Crowe, Merrill, Neas, Thompson, Guthridge, James.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We would like to thank all of the study participants for their time and commitment to the study, as well as their referring physicians, Drs. C. Carson, A. A. Kumar, L. Zacharias, J. B. Harley, and physician assistants T. Aberle and J. K. Shoemaker. We would like to thank the laboratory of Gillian M. Air, PhD (University of Oklahoma Health Sciences Center) for the determinations of humoral responses to influenza vaccination. We would also like to thank J. Anderson, W. Klein, G. Vidal, W. DeJager, B. Faris, and J. Levin for technical assistance and Scott Stewart for assistance with statistical analysis.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES