The Role of CXCL10 and IL-18 as Markers of Repigmentation Response in Nonsegmental Vitiligo Treated with Narrowband UVB Phototherapy: A Prospective Cohort Study

Vitiligo is a depigmenting disease caused by the loss of melanocytes owing to autoimmune cytotoxic activity and, as was recently described, basal melanocyte detachment (Boukhedouni et al., 2020Boukhedouni N. Martins C. Darrigade A.S. Drullion C. Rambert J. Barrault C. et al.Type-1 cytokines regulate matrix metalloprotease-9 production and E-cadherin disruption to promote melanocyte loss in vitiligo.JCI Insight. 2020; 5e133772PubMed Google Scholar). IFN-γ plays an important role in CXCL10 secretion and participates in TCD8+ cell recruitment to the affected skin through its receptor, CXCR3 (Wang et al., 2016Wang X.X. Wang Q.Q. Wu J.Q. Jiang M. Chen L. Zhang C.F. et al.Increased expression of CXCR3 and its ligands in patients with vitiligo and CXCL10 as a potential clinical marker for vitiligo.Br J Dermatol. 2016; 174: 1318-1326Crossref PubMed Scopus (91) Google Scholar). CXCL10 is significantly elevated in more severe vitiligo (Abdallah et al., 2018Abdallah M. El-Mofty M. Anbar T. Rasheed H. Esmat S. Al-Tawdy A. et al.CXCL-10 and interleukin-6 are reliable serum markers for vitiligo activity: a multicenter cross-sectional study.Pigment Cell Melanoma Res. 2018; 31: 330-336Crossref PubMed Scopus (29) Google Scholar; Zhang et al., 2020Zhang L. Chen S. Kang Y. Wang X. Yan F. Jiang M. et al.Association of clinical markers with disease progression in patients with vitiligo from China.JAMA Dermatol. 2020; 156: 288-295Crossref PubMed Scopus (10) Google Scholar) and is important in the maintenance of the depigmented phenotype in a mouse model of vitiligo (Rashighi et al., 2014Rashighi M. Agarwal P. Richmond J.M. Harris T.H. Dresser K. Su M.W. et al.CXCL10 is critical for the progression and maintenance of depigmentation in a mouse model of vitiligo.Sci Transl Med. 2014; 6: 223ra23Crossref PubMed Scopus (260) Google Scholar). Cytokine IL-18 promotes IFN-γ induction and TCD8+ cell activity and is proposed to act through NLRP3 inflammasome pathway activation in vitiligo (Tang and Zhou, 2020Tang L. Zhou F. Inflammasomes in common immune-related skin diseases.Front Immunol. 2020; 11: 882Crossref PubMed Scopus (20) Google Scholar). We recently demonstrated that patients with nonsegmental vitiligo have different response velocities during narrowband UVB phototherapy (Cabrera et al., 2018Cabrera R. Hojman L. Recule F. Sepulveda R. Delgado I. Predictive model for response rate to narrowband ultraviolet B phototherapy in vitiligo: a retrospective cohort study of 579 patients.Acta Derm Venereol. 2018; 98: 416-420Crossref PubMed Scopus (5) Google Scholar). Nevertheless, we do not know how these inflammatory cytokines and cytotoxic cells behave during treatment. After approval by the institutional review board, we recruited and obtained written informed consent from 23 patients from the Phototherapy Unit of Clinica Alemana de Santiago (Vitacura, Santiago, Chile) between April 2016 and June 2018 (Supplementary Table S1) and followed them during 48 sessions (three times weekly) of phototherapy. Every patient was prescribed phototherapy because they were classified as having a progressive disease (defined as the appearance of new achromic patches or enlargement of previous ones in the 3 months preceding enrollment); in addition, all patients were held off from other treatments during a minimum of 1 month before enrollment. Six patients were classified as slow responders compared with 17 patients who were classified as fast responders (Supplementary Table S2). The slow responders had significantly more time from diagnosis (Supplementary Table S2). Most patients improved Vitiligo Area Scoring Index (VASI) score after 24 sessions of phototherapy, with a mean reduction of 42%. Interestingly, mainly fast responders contributed to VASI reduction (Figure 1a). In contrast, patients with less than 2 years of vitiligo diagnosis, classified as having recent-onset vitiligo, had significantly higher VASI improvement than patients with longstanding vitiligo (Figure 1b). Time from diagnosis showed a positive correlation with VASI at baseline and a negative correlation with VASI improvement at 24 sessions (Figure 1c and d). Blood samples were taken at 24- and 48-session checkpoints, and flow cytometry was performed to select live CD3+CD4‒CD8+ cells (Supplementary Materials and Methods) (Figure 1e). Circulating TCD8+ cells decreased during phototherapy, reaching a mean of 45% at 48 sessions (Figure 1f). In addition, TCD8+ cells correlated indirectly with VASI improvement at 24 sessions and directly with CXCL10 (Figure 1g and h), showing that when the patient shows a better response to narrowband UVB phototherapy, circulating cytotoxic T cells and serum CXCL10 diminish. Considering this and knowing that CXCL10 and IL-18 participate in the inflammatory lymphocyte response in the skin, we aimed to measure serum CXCL10 and IL-18 levels at baseline and at 24 and 48 sessions of phototherapy. IL-18 decreased significantly at 24 and 48 sessions with a mean of 45% (Figure 2a). When analyzing IL-18 in subgroups, only fast responders had decreased IL-18 after 24 and 48 sessions (Figure 2b). CXCL10, in addition, diminished significantly at 24 and 48 sessions, paralleling the behavior of IL-18 (Figure 2c). Fast responders, as in IL-18 measurements, showed lower CXCL10 levels after treatment, with a mean of 42% at 24 sessions (Figure 2d). When divided into recent-onset and longstanding vitiligo, those with recent-onset vitiligo had higher CXCL10 levels at baseline, and these CXCL10 levels were the only ones that diminished after phototherapy (Figure 2e), correlating positively and significantly with IL-18 after 24 sessions (Figure 2f). Finally, to clarify the IL-18 possible role in INF-γ production, we analyzed the CD3+CD4‒CD8+INF-γ+ cell count showing a significant decrease at 48 sessions (Figure 2g). In addition, when plotted against IL-18 levels, INF-γ+ cells correlated positively and significantly (Figure 2h). These results support the role of CXCL10 and IL-18 in cytotoxic activity in vitiligo. It has been described that the activation of the IFN-γ, CXCL10, and CXCR3 axis is important for triggering and maintaining vitiligo in a mouse model (Frisoli et al., 2020Frisoli M.L. Essien K. Harris J.E. Vitiligo: mechanisms of pathogenesis and treatment.Annu Rev Immunol. 2020; 38: 621-648Crossref PubMed Scopus (59) Google Scholar); in humans, CXCL10 has been reported both as a severity marker (Abdallah et al., 2018Abdallah M. El-Mofty M. Anbar T. Rasheed H. Esmat S. Al-Tawdy A. et al.CXCL-10 and interleukin-6 are reliable serum markers for vitiligo activity: a multicenter cross-sectional study.Pigment Cell Melanoma Res. 2018; 31: 330-336Crossref PubMed Scopus (29) Google Scholar) and as a potential therapeutic target (Rodrigues et al., 2017Rodrigues M. Ezzedine K. Hamzavi I. Pandya A.G. Harris J.E. Vitiligo Working GroupNew discoveries in the pathogenesis and classification of vitiligo.J Am Acad Dermatol. 2017; v77: 1-13Abstract Full Text Full Text PDF Scopus (156) Google Scholar). In contrast, IL-18 has not been studied for vitiligo until now. Cytotoxic activity against melanocytes by TCD8+ cells causes its elimination in vitiligo lesions (Frisoli et al., 2020Frisoli M.L. Essien K. Harris J.E. Vitiligo: mechanisms of pathogenesis and treatment.Annu Rev Immunol. 2020; 38: 621-648Crossref PubMed Scopus (59) Google Scholar), and the roles of many cytokines have been demonstrated in its pathogenesis (IL-2, IL-6, IL-8, GM-CSF, TNF-α, IFN-γ, IL-4, IL-17, IL-10, IL-13, TGF-β1) (Tembhre et al., 2013Tembhre M.K. Sharma V.K. Sharma A. Chattopadhyay P. Gupta S. T helper and regulatory T cell cytokine profile in active, stable and narrow band ultraviolet B treated generalized vitiligo.Clin Chim Acta. 2013; 424: 27-32Crossref PubMed Scopus (52) Google Scholar; Yu et al., 1997Yu H.S. Chang K.L. Yu C.L. Li H.F. Wu M.T. Wu C.S. et al.Alterations in IL-6, IL-8, GM-CSF, TNF-alpha, and IFN-gamma release by peripheral mononuclear cells in patients with active vitiligo.J Invest Dermatol. 1997; 108: 527-529Abstract Full Text PDF PubMed Scopus (88) Google Scholar). We published previously different repigmentation rates in vitiligo treated with narrowband UVB. We found individuals who responded very rapidly to phototherapy (very fast and fast responders), others who responded very slowly (slow and very slow responders), and a third group whose response was in between these two subgroups (average responders). Although the majority of the fast responders had facial vitiligo, some of them also had vitiligo only in the trunk and extremities (Cabrera et al., 2018Cabrera R. Hojman L. Recule F. Sepulveda R. Delgado I. Predictive model for response rate to narrowband ultraviolet B phototherapy in vitiligo: a retrospective cohort study of 579 patients.Acta Derm Venereol. 2018; 98: 416-420Crossref PubMed Scopus (5) Google Scholar). To apply this velocity repigmentation rate classification, we got results confirming that fast responders—although had higher baseline levels of CXCL10 and IL-18—under narrowband UVB treatment, can lower blood levels of these cytokines at a higher rate than that of slow responders. In our cohort, despite that all patients had progressive vitiligo, those with the longstanding disease had a lower rate of response to narrowband UVB phototherapy. They also had a lower baseline concentration of CXCL10 and IL-18. These findings suggest that patients with recent-onset vitiligo have a better response to phototherapy than those with longstanding vitiligo and that CXCL10 and IL-18 can probably be used as biologic markers to predict and monitor the clinical response to narrowband UVB phototherapy in patients with nonsegmental vitiligo. Further studies are needed to confirm this hypothesis. All data generated or analyzed during this study are included in this published article and its supplementary information. Lía Hojman: http://orcid.org/0000-0003-0501-4156 Raúl Cabrera: http://orcid.org/0000-0002-0180-9130 Claudio Karsulovic: http://orcid.org/0000-0002-8819-8546 Fabian Tempio: http://orcid.org/0000-0002-2655-6061 Claudio Perez: http://orcid.org/0000-0002-5983-6596 Mercedes López: http://orcid.org/0000-0002-1744-2038 The authors state no conflict of interest. Conceptualization: LH, RC, ML, CK; Data Curation: LH, CK; Formal Analysis: LH, RC, ML, CK; Funding Acquisition: LH; Investigation: LH, ML, FT, CK, CP, RC; Resources: LH, ML; Supervision: RC, ML; Writing - Original Draft Preparation: LH, RC, CK; Writing - Review and Editing: LH, RC, CK Inclusion criteria for patients are described in Supplementary Table S1. Disease diagnosis was assessed independently by two dermatologists (RC and LH) using daylight and Wood's light (VISIA system 4th generation; Canfield Industries, Youngstown, OH). Vitiligo Area Scoring Index (VASI) score was measured at 0, 24, and 48 sessions with standardized photographs (Canon camera EOS t3i, Tokyo, Japan). CXCL10 and IL-18 levels were measured using ELISA (ELISA kit; anti-CXCL10 and anti-IL-18; BD Biosciences, San Jose, CA). Flow cytometry (BD FACSverse flow cytometer; BD Biosciences) was used for the measurement of TCD8+ cells (live and/or dead; anti-CD3, anti-CD4, anti-CD8; BD Biosciences). Cytokine levels and TCD8+ cells were measured at the same checkpoints as VASI. VASI improvement was calculated as ([baseline VASI ‒ VASI]/baseline VASI) × 100. We classified patients as slow or fast responders using a modification of our predictive model of response for patients with vitiligo (Cabrera et al., 2018Cabrera R. Hojman L. Recule F. Sepulveda R. Delgado I. Predictive model for response rate to narrowband ultraviolet B phototherapy in vitiligo: a retrospective cohort study of 579 patients.Acta Derm Venereol. 2018; 98: 416-420Crossref PubMed Scopus (7) Google Scholar): patients who showed ≥10% repigmentation at 24 sessions and/or ≥25% repigmentation at 48 sessions were considered fast responders. Those who showed less repigmentation were considered slow responders. We divided patients into those with recent-onset vitiligo (<24 months from diagnosis) and those with longstanding vitiligo (>24 months). Patients received narrowband UVB phototherapy three times a week using the standardized protocol previously published (Mohammad et al., 2017Mohammad T.F. Al-Jamal M. Hamzavi I.H. Harris J.E. Leone G. Cabrera R. et al.The Vitiligo Working Group recommendations for narrowband ultraviolet B light phototherapy treatment of vitiligo.J Am Acad Dermatol. 2017; 76: 879-888Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar). All patients received 24 narrowband UVB sessions, and 10 patients (43.5%) received 48 narrowband UVB sessions to achieve a better repigmentation. We used GraphPad software and Stata 13.1 (GraphPad Software, San Diego, CA) for data analysis. Continuous variables were expressed as means (SDs), and discrete variables were expressed as percentage distributions. Fisher exact and chi-square tests were used to compare qualitative data, and t-test and ANOVA test were used for quantitative data. Regression models were constructed according to the distribution and nature of the variables. Correlation tests were performed using Spearman correlation tests. P < 0.05 was considered to be statistically significant.Supplementary Table S1Inclusion and Exclusion Criteria for the StudyInclusionExclusionAged >14 yPersonal history of malignant melanomaClinical or histological diagnosis of nonsegmental vitiligoUntreated or currently under treatment for cancerWithout current history or past 30‒day history of vitiligo treatmentHistory of primary or secondary immunodeficiencySigned written informed consentSubjects with a history of photosensitive pathologies such as porphyria, lupus, and untreated skin cancer—ANA titers >1/80—Patients unable to stay in the standing positionHistory of claustrophobiaAbbreviation: ANA, antinuclear antibodies. Open table in a new tab Supplementary Table S2Baseline Demographic Characteristics of the ParticipantsCharacteristicsAll Patients(N = 23)Patients with Recent-Onset Vitiligo1Recent-onset vitiligo includes vitiligo diagnosis <24 months.(n = 9)Patients with Longstanding Vitiligo2Longstanding vitiligo includes vitiligo diagnosis >24 months.(n = 14)Slow Responders(n = 6)Fast Responders(n = 17)P-ValueP-ValueAge, y, mean (SD)35.5 (10.6)34.1 (11.2)37.5 (11.5)36.7 (13)35.5 (9.9)0.520.40Female, n (%)10 (43)2 (22)8 (57)5 (83)5 (29.4)0.070.64Time since diagnosis, mo, mean (SD)104.8 (141)9.6 (10.2)178 (151.6)224.6 (173.3)64.7 (106.2)0.00020.00081 Recent-onset vitiligo includes vitiligo diagnosis <24 months.2 Longstanding vitiligo includes vitiligo diagnosis >24 months. Open table in a new tab Abbreviation: ANA, antinuclear antibodies.