Clinical Significance of Serum Soluble CD Molecules to Assess Disease Activity in Vitiligo

Key Points

Question  Are serum soluble CD (sCD) levels associated with disease activity in vitiligo?

Findings  In this cross-sectional study of 93 patients with vitiligo, serum sCD25 and sCD27 were significantly linked with disease activity in vitiligo. Moreover, sCD27 was associated with a capacity to indicate probable future disease progression.

Meaning  This study supports the role of sCD25 and sCD27 as biomarkers to assess vitiligo activity and indicate likely future progression.

Importance  It is difficult to determine disease activity in vitiligo owing to the absence of inflammatory signs, such as erythema or scaling. A biomarker that could confirm active disease and indicate likely future disease progression would therefore be of considerable value.

Objective  To investigate whether soluble CD27 (sCD27), sCD25, or sCD40L could be valuable biomarkers to determine disease activity in vitiligo and indicate likely future progression.

Design, Setting, and Participants  A combined cross-sectional and prospective study was conducted at the department of dermatology at Ghent University Hospital between February 24, 2012, and December 12, 2015. Ninety-three patients with vitiligo were enrolled, including 83 individuals with nonsegmental vitiligo and 10 with segmental vitiligo. Blood sampling was performed, and sCD25, sCD27, and sCD40L were measured in serum.

Main Outcomes and Measures  The associations between sCD levels, disease activity, and future progression were investigated.

Results  Of the 93 patients included in the study, 51 were women (55%); median (interquartile range) age was 36.5 (26.0-49.8) years. Both sCD27 (21.5 ng/mL [16.1-30.0 ng/mL] vs 18.4 ng/mL [12.5-22.1 ng/mL]; P = .006) and sCD25 (2.6 ng/mL [2.1-3.4 ng/mL] vs 2.2 ng/mL [1.7-2.4 ng/mL]; P = .002) levels were associated with active disease. Moreover, a statistically significant link with disease progression after 3 to 6 months was found for sCD27 (21.7 [17.0-29.1] vs 16.6 [13.5-23.7]; P = .02) but not for sCD25 (2.8 ng/mL [2.2-3.4 ng/mL] vs 2.3 [1.9-2.8 ng/mL]; P = .053). Further in vitro experiments showed a correlation between sCD25 and interferon γ (r = 0.562, P = .005), interleukin 10 (r = 0.453, P = .03), and sCD27 secretion (r = 0.549, P = .007). No associations were found for sCD40L levels.

Conclusions and Relevance  This study demonstrates increased levels of sCD27 and sCD25 in patients with active vitiligo. Moreover, these results provide the first evidence that these markers have a capacity to indicate the probability of future disease progression, which supports their role as biomarkers in vitiligo.

Limited evidence is available on circulating markers that are linked to progressive vitiligo. Nonetheless, because an increased extent of vitiligo is associated with impaired quality of life, adequate identification of disease activity is crucial for therapeutic management.¹ Several associations between immunologic markers and vitiligo activity have been reported² investigating the protein, RNA, and microRNA levels. Most studies are hampered by limited patient size and lack prospective follow-up. The link between soluble CD (sCD) molecules and several autoimmune disorders is well established.³ Soluble CD molecules are shed from activated and proliferating lymphocytes, which makes them useful markers to confirm immunologic activity.

Signaling of CD25 (also termed IL-2 receptor) is a critical factor in maintaining T-cell tolerance, especially regarding regulatory T-cell development.⁴ In general, sCD25 is relatively stable over time, which makes it an attractive target to study inflammatory disorders. Conflicting evidence exists as to whether sCD25 is a marker of disease activity in vitiligo.^5-7 Therefore, additional research is needed on this topic, especially considering the lack of prospective follow-up in all previously performed studies.

A transmembrane protein, CD27 is expressed on T, B, and natural killer cells and belongs to the tumor necrosis factor receptor family. With binding to its ligand (CD70), CD27 ensures lymphocyte survival, increased T-cell proliferation, and memory cell formation. Its expression supports helper T-cell type 1 (T_h1) development. Soluble CD27 is released from the cell surface of activated lymphocytes by shedding that is induced by metalloproteinases or by differential splicing of the receptor protein. Increased levels have been found⁸ in systemic lupus erythematosus, celiac disease, viral infections, and lymphoid cancers, although, to our knowledge, it has never been studied in vitiligo.

The CD40-CD40L pathway is involved in multiple autoimmune disorders, such as thyroiditis, psoriasis, rheumatoid arthritis, Sjögren syndrome, and systemic lupus erythematosus. In addition, CD40L is expressed on a variety of immune cells and, after CD40-CD40L ligation, induces activation of T and B lymphocytes. Furthermore, CD40-CD40L interaction leads to maturation of dendritic cells toward an antigen-presenting phenotype.^9,10 The purpose of this study was to investigate whether sCD27, sCD25, and sCD40L could be useful biomarkers to determine disease activity in vitiligo and indicate probable future progression.

Ninety-three patients with vitiligo were enrolled in this combined, cross-sectional, prospective study at the Department of Dermatology at Ghent University Hospital, Ghent, Belgium, between February 24, 2012, and December 12, 2015. Eighty-three patients with nonsegmental vitiligo and 10 individuals with segmental vitiligo were included; 51 women (55%) and 42 men (45%) with a median age of 39.5 years (interquartile range [IQR], 27.0-52.8) participated in this study. The mean affected body surface area was 4.47% (median, 1.5% [IQR, 0.5%-5.0%]). The participants’ clinical characteristics are summarized in Table 1. Patients receiving no therapy or applying topical formulations (excluding UV-B treatment) were enrolled. Patients who had received no treatment during 3 months before blood analysis were considered as untreated. The study was approved by the ethics committee of Ghent University Hospital. All patients signed written informed consent; no financial compensation was provided.

Patient-reported disease activity was evaluated by using a questionnaire assessing disease activity, stability, and repigmentation during the past 3, 6, or 12 months. Disease activity was assessed by the physician and classified into 4 categories: very active, moderate or low active, stable disease, and repigmentation. This categorization was accomplished using digital follow-up photographs taken with or without a Wood lamp and clinical examination. Furthermore, subsequent disease evolution was investigated during a follow-up consultation after 6 to 12 months.

Blood sampling was performed, and sCD25, sCD27, and sCD40L (R&D Systems) were measured in serum using a commercial kit (Human Fluorokine MAP Base Kit; R&D Systems). Further in vitro experiments were performed to investigate the association between soluble CD release and cytokine production. Briefly, peripheral blood mononuclear cells obtained from 10 patients with vitiligo and 10 healthy individuals serving as controls were stimulated with anti-CD3/CD28 beads; supernatant was harvested after 3 days for measuring sCD25 and sCD27.

For comparison between nonparametric variables, the Mann-Whitney test or Kruskal-Wallis test was used. P < .05 was considered to indicate statistical significance. Adjustment for possible confounding factors (use of treatment) was performed using multivariate linear regression models in case of a continuous dependent variable and logistic regression in case of a dichotomous dependent variable. The Wald statistic in logistic regression was also calculated. For correlation analysis, Pearson correlation was conducted. Receiver operating characteristic (ROC) analysis was performed to determine the sensitivity and specificity of sCD25 and sCD27 and the area under the ROC curve. All statistical analyses were performed using SPSS, version 23.0 (SPSS Science).

Regarding patient-reported disease activity, elevated sCD27 levels were significantly linked with disease activity in the past year (median [IQR], 21.5 ng/mL [16.1-30.0 ng/mL] vs 18.4 ng/mL [12.5-22.1 ng/mL]; P = .006) (Figure 1B). Accordingly, higher sCD27 values were found in patients with active vitiligo (20.6 ng/mL [15.8-30.2 ng/mL] vs 18.4 ng/mL [13.6-24.2 ng/mL]; P = .045) in which the disease activity was scored by the physician. In patients who showed clear repigmentation, sCD27 values were markedly lower (P = .008) compared with all other participants with vitiligo. However, no significant difference was found between patients who reported recent disease activity compared with those whose disease had progressed during the past year but had developed no new lesions in the past months (Figure 1C). Patients with segmental vitiligo displayed sCD27 values comparable to those seen in patients with stable generalized vitiligo.

Similar results were observed for sCD25. Higher values were found in patients who reported disease activity in the past year (median [IQR], 2.6 ng/mL [2.1-3.4 ng/mL] vs 2.2 ng/mL [1.7-2.4 ng/mL]; P = .002) (Figure 1A) and in patients with active disease confirmed by clinical follow-up photographs (2.6 ng/mL [2.1-3.4 ng/mL] vs 2.2 ng/mL [1.8-2.7 ng/mL]; P = .04).

The sCD40L levels were also higher in patients with active disease, although this elevation was less pronounced compared with the other investigated CD molecules. Nonetheless, patients with active vitiligo had significantly more CD40L values exceeding 20 ng/mL (53/87 [60.9%] vs 34/87 [39.1%]; P = .008) compared with levels in patients with stable and repigmenting vitiligo.

Patients applying topical therapy had decreased sCD25 values (mean [IQR], 2.4 ng/mL [2.0-3.4 ng/mL] vs 2.2 ng/mL [1.8-2.6 ng/mL]; P = .03), which may be attributed to treatment-induced decreased disease activity. However, sCD25 values were lower in both the active and stable vitiligo groups receiving topical treatment (Figure 2A). In contrast, sCD27 and sCD40L levels were not influenced by the administration of topical therapy. In multivariate logistic regression models, both sCD27 and sCD25 remained significantly associated with disease activity in the past year after adjustment for local therapy (Table 2).

The ROC curve analysis for assessing disease activity was significant for sCD27 and sCD25, displaying areas under the curve of 0.694 and 0.716, respectively. For sCD27, a cutoff value of 25 ng/mL resulted in a specificity of 88.5% and a sensitivity of 43.3%. For sCD25, a cutoff set at 2.5 ng/mL displayed a specificity of 87.5% and a sensitivity of 50.0% (Figure 2B).

Soluble CD molecules were not significantly linked to the affected body surface area; sCD27 and sCD25 concentrations were highly correlated (P < .001; ρ = 0.51) (Figure 1E). None of the soluble CD markers were associated with sex, age, disease duration, or the presence of halo nevi. There was also no link found in participants with a personal history of or first-degree relatives with autoimmune diseases.

Data on prospective follow-up were available in 64 patients. Soluble CD27 was significantly associated with future disease progression (median [IQR], 21.7 [17.0-29.1] vs 16.6 [13.5-23.7]; P = .02), and sCD25 was elevated, but not significantly so (2.8 ng/mL [2.2-3.4 ng/mL] vs 2.3 [1.9-2.8 ng/mL]; P = .053), in the group of patients with disease progression (Figure 2C and 2D). In a logistic regression model, the value of sCD27 as an indicator of future vitiligo evolution was confirmed after adjustment for the use of treatment during follow-up (odds ratio, 1.10 [95% CI, 1.01-1.19]; P = .04).

Further in vitro experiments were performed to investigate the association between sCD release and cytokine production. Enhanced sCD25 levels were found in patients with vitiligo (n = 10) compared with healthy controls (n = 10), but no difference was found in sCD27 secretion. Additional experiments showed that sCD25 release was associated with IFN-γ (r = 0.562, P = .005) and IL-10 (r = 0.453, P = .03) production (Figure 3). Similar to the results found in serum, sCD25 and sCD27 levels in the supernatant were correlated (r = 0.549, P = .007). In the supernatant, the concentrations of sCD27 were lower than those of sCD25; however, sCD27 values in serum were approximately 10-fold higher than those of sCD25 (median [IQR] for sCD27, 19.1 [15.5-26.8] vs sCD25, 2.3 [1.9-2.9]).

This study demonstrates higher values of sCD27 and sCD25 in the serum of patients with active vitiligo and, to our knowledge, points for the first time to their beneficial use as indicators of disease progression. Despite the fact that sCD markers have been recognized for several decades, their immunologic role has only recently been clarified. Instead of being a bystander product from activated lymphocytes, the active contribution of sCD markers to inflammatory processes has recently been confirmed. These contributions include the immunologic-stimulating effects of sCD27 by inducing T-cell activation.⁸ In addition, sCD27 increases IgG production and provides an activation signal for antigen-primed B lymphocytes,¹¹ and sCD25 has been implicated in the development of autoimmunity. Through inhibiting the downstream signaling of IL-2R, sCD25 favors T_H17 development. By acting as a decoy receptor for IL-2, T-cell responses are skewed toward a T_H17 phenotype.⁴ In this regard, increased circulating T_H17 cells and increased serum levels of IL-17 have been associated with active vitiligo.^12,13

The correlation between sCD25 and proinflammatory cytokine production (especially interferon γ [IFN-γ]) in our in vitro experiments further corroborates the association of sCD25 with an activated immune state. Moreover, this association fits with the current understanding of vitiligo pathogenesis since gene expression profiling has revealed elevated expression of IFN-γ and associated genes in vitiligo skin.¹⁴ Cytotoxic T cells in the skin and blood of patients with vitiligo produce IFN-γ, and IFN-γ production of circulating T cells has been shown¹⁵ to correlate with disease activity. In vitro experiments¹⁶ showed that IFN-γ has direct cytotoxic effects on melanocytes. Moreover, CXCL10, which is an IFN-γ–induced chemokine, is elevated in the serum of patients and functions as a chemoattractant for activated T cells. Neutralization of IFN-γ and CXCL10 was able to reverse depigmentation in mouse models for vitiligo.¹⁷ In addition, mouse experiments¹⁸ demonstrated that CXCL10 is the exclusive chemokine released by primed cytotoxic T cells after CD27-CD70 costimulation. Two cases^19,20 have been described of beneficial responses in patients with vitiligo after JAK inhibition, which supports the key role of the IFN-γ pathway in vitiligo since JAK1 is required for IFN-γ signaling.^19-21 The correlation between sCD25 and IFN-γ in our in vitro experiments supports the hypothesis that soluble CD molecules reflect a key aspect of vitiligo pathogenesis and activity. Furthermore, the decrease of sCD25 values in patients with vitiligo receiving topical treatment suggests that we might be able to counter this phenomenon therapeutically.

In this study, sCD27 and, to a lesser extent, sCD25 had a predictive capacity on future progression of the vitiligo lesions. This finding supports the possible role of these molecules as biomarkers in clinical practice. The fact that these markers reflect immunologic activation is appealing since most current treatments for vitiligo act through immunosuppressive mechanisms. Although considerable overlap exists between patients with progressive and nonprogressive vitiligo, these results, based on a model with only 1 predictive marker, are promising. With use of cutoff values for circulating CD molecules, approximately half of the patients with active vitiligo can be identified with an acceptable false-positive rate of approximately 12%. Future studies with consecutive blood sampling are likely to display even better results by taking into account individual differences in baseline sCD27 and sCD25 production. A limitation of this study is the monocentric design with a relatively small patient size, which hampers the interpretation of the ROC analyses.

This study shows that sCD27 and sCD25 values in the serum of patients with vitiligo are associated with disease activity. These markers also have a capacity to indicate future disease progression. Future studies with consecutive blood sampling may determine their final usefulness as biomarkers in clinical practice.

Corresponding Author: Reinhart Speeckaert, MD, PhD, Department of Dermatology, Ghent University Hospital, De Pintelaan 185, 9000 Ghent, Belgium (reinhart.speeckaert@ugent.be).

Accepted for Publication: May 28, 2016.

Published Online: August 24, 2016. doi:10.1001/jamadermatol.2016.2366.

Author Contributions: Drs Speeckaert and van Geel contributed equally to the study, had full access to all the data in the study, and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Speeckaert, van Geel.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Speeckaert.

Critical revision of the manuscript for important intellectual content: All authors.

Obtained funding: Speeckaert, van Geel.

Administrative, technical, or material support: Speeckaert, van Geel.

Study supervision: All authors.

Conflict of Interest Disclosures: None reported.

Funding/Support: This research was supported by postdoctoral research grant 01P12914 (Dr Speeckaert) from the Ghent University Special Research Fund and research grant 1831512N (Dr van Geel) from the Scientific Research Foundation–Flanders.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.