Randomized, Double-Blind, Placebo-Controlled Trial of the Effect of Vitamin D3 on the Interferon Signature in Patients With Systemic Lupus Erythematosus

Abstract

Objective: Vitamin D modulates the immune response and blocks induction of an interferon (IFN) signature by systemic lupus erythematosus (SLE) sera. This study was undertaken to investigate the effects of vitamin D supplementation on the IFN signature in patients with SLE.

Methods: SLE patients (n = 57) with stable, inactive disease, a serum 25-hydroxyvitamin D (25[OH]D) level ≤20 ng/ml, an elevated anti-double-stranded DNA antibody level, and an IFN signature (as determined by measuring the expression levels of 3 IFN response genes) were randomized into a 12-week double-blind, placebo-controlled trial of vitamin D3 at doses of 2,000 IU or 4,000 IU. An IFN signature response was defined as a 50% reduction in the expression of 1 of the 3 genes or a 25% reduction in the expression of 2 of the 3 genes. Disease activity, adverse events, and endocrine effects were assessed.

Results: Baseline characteristics of the patients in the 3 treatment groups (placebo, low-dose vitamin D3 , or high-dose vitamin D3 ) were similar. Repletion of 25(OH)D (i.e., levels ≥30 ng/ml) was not observed in any of the patients who were receiving placebo, while repletion was observed in 16 of 33 patients receiving vitamin D3 . The percentage of patients with an IFN signature response did not differ among the treatment groups. Moreover, there was no difference in the percentage of patients with an IFN signature response between those who remained vitamin D deficient and those who demonstrated repletion of vitamin D. Modular microarray analysis of a subset of patients (n = 40) did not reveal changes from baseline in any modules (including the IFN-inducible module) in any of the treatment groups, and no differences in expression were found between patients who demonstrated vitamin D repletion and patients who were persistently vitamin D deficient. Vitamin D3 was well tolerated, and there were no safety concerns.

Conclusion: Vitamin D3 supplementation up to 4,000 IU daily was safe and well-tolerated but failed to diminish the IFN signature in vitamin D-deficient SLE patients. Higher 25(OH)D levels sustained for a longer duration may be required to affect immunologic outcomes.

Trial registration: ClinicalTrials.gov NCT00710021.

© 2015, American College of Rheumatology.

Figures

Figure 1

125 subjects were screened for eligibility into this study with randomization of 57 subjects (19 per treatment arm). The modified intent-to-treat (mITT) population was comprised of 54 subjects as 3 subjects never received a dose of study medication. The per-protocol (PP) population consisted of 31 subjects who had an overall treatment compliance 80-120% and had no substantive protocol deviations. (Subjects were excluded from the PP population if they had at least one of the following: noncompliant (n=6), negative IFN signature at baseline (n=7), receipt of prohibited medications at baseline (n=3), increase of immunosuppression during course of study (n=5), or, missing week 12 vitamin D (n=2) or IFN signature data (n=5)).

Figure 2

Serum 25(OH)D levels over time by treatment group. Mean serum 25(OH)D values and SD are shown for baseline, Week 6 and Week 12 study visits for each treatment group as well as for the pooled (low and high dose) group. mITT population with available data.

Figure 3

Scatter plots of M×1 (A), Ifit1 (B), and Ifi44 (C) log2 ratios of Week 12 to baseline gene expression versus serum 25(OH)D at Week 12 are displayed. Changes in gene expression in responders (open symbols) and non-responders (closed symbols) are shown for subjects in each treatment group.

Figure 4

Change in expression of each gene from baseline. Mean (95% CI) for gene expression ratio (Week 12: baseline), on the log2 scale, after adjustment for baseline gene expression. Log2 of the gene expression ratio equals, if positive, the number of time expression doubled in 12 weeks, if negative, the number of times expression was halved. P-values are for the overall treatment effect (df=2) after adjustment for baseline expression levels (ANCOVA).