ORIGINAL RESEARCH published: 25 September 2020 doi: 10.3389/fimmu.2020.579005 Systemic Regulatory T Cells and IL-6 as Prognostic Factors for Anatomical Improvement of Uveitic Macular Edema Jessica Matas 1, Victor Llorenç 1, Alex Fonollosa 2, David Dı́az-Valle 3, Cristina Esquinas 4, Maria Teresa Sainz de la Maza 1, Marc Figueras-Roca 1, Joseba Artaraz 2, Barbara Berasategui 2, Marina Mesquida 1, Alfredo Adán 1 and Blanca Molins 1* 1Group of Ocular Inflammation, Institut d’Investigacions Biomèdiques Agustı´ Pi i Sunyer (IDIBAPS), Hospital Clı´nic de Barcelona, Barcelona, Spain, 2 Department of Ophthalmology, BioCruces Health Research Institute, Hospital Cruces, University of the Basque Country, Baracaldo, Spain, 3 Ophthalmology Department and Health Research Institute (IdISSC), Edited by: Hospital Clinic of San Carlos, Madrid, Spain, 4 Valle Hebron Research Institute, Autonomous University of Barcelona, Peizeng Yang, Barcelona, Spain First Affiliated Hospital of Chongqing Medical University, China Purpose: To investigate whether systemic immune mediators and circulating regulatory T Reviewed by: Yohei Tomita, cells (Tregs) could be prognostic factors for anatomic outcomes in macular edema Boston Children’s Hospital and secondary to non-infectious uveitis (UME). Harvard Medical School, United States Kuender D. Yang, Methods: Multicenter, prospective, observational, 12-month follow-up study of 60 Mackay Memorial Hospital, Taiwan patients with UME. Macular edema was defined as central subfield thickness (CST) > *Correspondence: 300 mm measured with spectral domain optical coherence tomography (SD-OCT). Serum Blanca Molins samples and peripheral blood mononuclear cells (PBMC) were obtained from venous bmolins@clinic.cat blood extraction at baseline. Serum levels of IL-1b, IL-6, IL-8, IL-17, MCP-1, TNF-a, IL-10, Specialty section: and VEGF were determined by Luminex. Tregs population, defined as CD3+CD4+FoxP3+ This article was submitted to in PBMC, was determined by flow cytometry. Main outcome measure was the predictive Autoimmune and Autoinflammatory Disorders, association between searched mediators and CST sustained improvement, defined as a section of the journal CST < 300 microns or a 20% CST decrease, at 6 months maintained until 12-months Frontiers in Immunology compared to baseline levels. Received: 01 July 2020 Accepted: 14 September 2020 Results: Multivariate logistic regression analysis showed an association between CST Published: 25 September 2020 sustained improvement at 12 months follow-up and IL-6 and Tregs baseline levels. Higher Citation: IL-6 levels were associated with less events of UME improvement (OR: 0.67, 95% CI Matas J, Llorenç V, Fonollosa A, Dı´az-Valle D, Esquinas C, (0.45–1.00), P = 0.042), whereas higher levels of Tregs favored such improvement (OR: de la Maza MTS, Figueras-Roca M, 1.25, 95% CI: 1.12–2.56, P = 0.049). Artaraz J, Berasategui B, Mesquida M, Adán A and Molins B (2020) Systemic Conclusions: Increased levels of Tregs and reduced levels of IL-6 in serum may be Regulatory T Cells and IL-6 prognostic factors of sustained anatomical improvement in UME. These findings could as Prognostic Factors for Anatomical Improvement of enforce the opportunity to develop more efficient and personalized therapeutic Uveitic Macular Edema. approaches to improve long-term visual prognosis in patients with UME. Front. Immunol. 11:579005. doi: 10.3389/fimmu.2020.579005 Keywords: macular edema, uveitis, cytokines, regulatory T cells, biomarkers Frontiers in Immunology | www.frontiersin.org 1 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME INTRODUCTION MATERIALS AND METHODS Non-infectious uveitis accounts for 10% of cases of total Study Design and Selection Criteria blindness in developed world and may result in persistent We conducted a multicenter, prospective, observational, 12- visual impairment in untreated patients (1). Uveitis may affect month follow-up study to analyze the association of peripheral people of all ages but most frequently occurs in the working age blood immune mediators and Tregs with the clinical evolution of population, thus implying a high socioeconomical burden (2). UME. Adult (>18 years-old) patients with non-infectious UME in Macular edema remains the most commonly found structural at least one eye were proposed for inclusion. UME was defined as complication of uveitis that results in central visual impairment central subfield thickness (CST) of >300 mm as measured by SD- although only one third of uveitis patients develop it (3). The OCT (HD-Cirrus, Carl Zeiss Meditec, Dublin, CA) and presence pathogenesis of uvetitic macular edema (UME) involves of fluid (intraretinal and/or subretinal) in the macula. Whenever breakdown of the blood-retinal barrier, followed by both bilateral UME was present, the eye with the higher CST value in intracellular and extracellular fluid accumulation within the the fovea according to OCT measurements in mm was selected as macular retina (4). However, the underlying pathophysiology the study eye. Exclusion criteria were limited to infectious uveitis, as well as the molecular mediators involved are not fully highly hazy ocular media, concurrent ocular diseases, exclusive understood, therefore making UME a controversial topic. tractional UME, pregnancy, immunocompromising systemic Chronic retinal inflammation has been observed in early diseases (including, but not limited to HIV, leukemia, phases and also in the sight-threatening advanced forms of lymphoma, and chemotherapy), and eyes with any intervention UME. The uveitic retina is characterized by a complex milieu (intraocular surgery, laser, and intravitreal therapy) in the of dysregulated proinflammatory factors. Increased levels of previous 4 months. certain proinflammatory cytokines, such as interleukin-6 (IL-6), Three referral centers for ocular inflammatory diseases in IL-8, and tumor necrosis factor-a (TNF-a), both in serum and Spain (Clı́nic Hospital of Barcelona, Cruces Hospital of Bilbao aqueous humor, have been found in patients with UME (5). and Clı́nico San Carlos Hospital of Madrid) participated in the Vascular endothelial growth factor (VEGF) may also play a major recruitment of patients from January 2015 to January 2017. Local role as it can be increased in aqueous humor of patients with UME ethics committees approved the study (Comité É tico de and in other forms of macular edema (5, 6). On the other hand, in Investigació n Clı́nica del Hospital Clı́nic de Barcelona 2013/ the last years, the interest in regulatory T cells (Tregs) as markers 8574; Comité de É tica de la Investigació n con Medicamentos de of activity and modulators of autoreactive T lymphocytes in Euskadi, Hospital Universitario Cruces PI201406; Comité É tico uveitis and other inflammatory diseases has gained attention de Investigació n Clı́nica del Hospital Clı́nico San Carlos de (7). Patients with active uveitis have reduced circulating levels of Madrid 13/244-E). All patients provided written informed Tregs compared to healthy subjects, with Tregs levels being consent, and the research followed the regulations of the significantly upregulated during quiescent disease (8, 9). Declaration of Helsinki. Moreover, adoptive transfer of Tregs has resulted in greater protection against intraocular inflammation in experimental Data Collection and Ophthalmic models of uveitis (10, 11). However, it is unknown whether Tregs Assessment are also involved in the pathophysiology of UME. Patients were evaluated at each clinical site, and protocol-based UME may persist despite adequate control of activity, thus assessments were performed at different time-points: baseline, leading to permanent photoreceptor damage and loss of central month 1, month 6, and month 12 of follow-up. Other visits at visual acuity (12). After the acute phase, UME may either subside different time-points were allowed at the discretion of the spontaneously or respond to treatment intended to reduce local treating physician. Data were recorded in an electronic case inflammatory mediators. Moreover, in relapsing and chronic report form designed ad hoc. forms of UME, especially in bilateral forms, systemic treatment Medical record data from each patient included demographics should be started, usually consisting of corticosteroids and/or (age, race, and sex), diagnosis classified by anatomic location immunomodulators (13). However, despite these options, a according to the Standardization of Uveitis Nomenclature number of patients still remain refractory to treatment and (SUN) criteria (14), laterality of disease, systemic or isolated UME may lead to severe visual impairment. ocular inflammatory disease classification, and a complete In summary, there is scarce data regarding the factors ophthalmic examination. influencing sustained anatomic and visual recovery in eyes Data gathered from evaluation of each patient included best- with UME. A better understanding of the prognostic corrected visual acuity (BCVA, with Snellen charts in decimal biomarkers in patients with UME would allow the use of more scale at a test distance of 6 m), the presence or absence of disease efficient and selective pharmacological approaches to improve activity, categorical and quantitative OCT findings, and UME long-term visual prognosis and minimizing unnecessary drug- treatment. The SUN recommendations were used for anterior induced toxicity. In this prospective study, we aimed to chamber and vitreous inflammatory activity gradations. SD- investigate the association of several systemic immune OCT exploration was used as determination of UME using mediators and Tregs with the anatomic outcome of UME after Macular cube 512 × 128 A-scan, within a 6 × 6 mm area 12 months of follow-up. centered on the fovea. Imaging assessment was performed by Frontiers in Immunology | www.frontiersin.org 2 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME two investigators who were masked to clinical data of the total of 50,000 events recorded for each sample through a live corresponding patients. In the event of discrepancies, the two CD3+CD4+ lymphocyte gate. Analysis of flow cytometry data graders made the assessment together and reached a consensus. was performed using FlowJo software. Tregs were defined as Masked investigators were asked to determine the CST and the CD3+CD4+FoxP3+ cells. pattern of UME (sub-retinal fluid, cystoid, diffuse, or tractional components). Macular volume, subfoveal choroidal thickness Outcome Measures (enhanced deep imaging mode), maximum diameter of the To determine whether Tregs and serum immune mediators can greatest cyst if present, subretinal fluid pocket’s length in a be predictive factors for anatomic outcomes in UME, we transfoveal B-scan were also recorded. Ancillary tests to rule analyzed the association between baseline levels of the above- out infectious origin (including but not limited to luetic mentioned immune mediators and Tregs with sustained serology, quantiFERON-TB Gold and toxoplasma serology) anatomical improvement of UME. Improvement of UME was and to appropriately classify the uveitis were ordered at the defined as a 20% CST decrease or CST < 300 µm at 6 months and investigator discretion. maintained at 12 months of follow-up compared to baseline. Sample Collection Statistical Analysis Two peripheral blood samples were obtained at the time of To describe the qualitative variables, absolute frequencies and patient inclusion in the presence of UME according to the percentages were used. The description of quantitative variables aforementioned definition. Serum was obtained from sample 1 was performed using the mean and standard deviation (SD). The as follows: blood was centrifugated at 1600 g within 20 min and Kolmogorov-Smirnov test was used to assess the normality of stored at -80°C until inflammatory mediators were determined. distributions. In the case of quantitative variables, the The levels of Tregs in peripheral blood were determined from comparison of the characteristics of the eyes depending on the sample 2: peripheral blood mononuclear cells (PBMCs) were presence of sustained anatomical improvement was carried out obtained by Ficoll (Ficoll-Plaque Plus, GE Healthcare) gradient using the Student t-test (or Mann-Whitney U-test if normality centrifugation, washed twice with RPMI 1640 containing 2% was not assumed). The Chi-squared test (Fisher test for heat-inactivated fetal calf serum and cryopreserved for further frequencies < 5) was used for the comparison of categorical staining and analysis by flow cytometry. Cryopreserved serum variables. Wilcoxon test was performed in order to analyze and PBMC samples were then shipped to the Ocular changes in CST during follow up. Inflammation Laboratory (Hospital Clinic of Barcelona) where A back stepwise logistic regression analysis was developed they were analyzed. including the defined and CST-based favorable anatomical outcome item as the dependent variable, and as independent Cytokine Determination variables, those cytokines and clinical variables with a P-value < Eight immune mediators were determined: IL-1b, IL-6, IL-8, IL- 0.2 in the univariate analysis and clinical pertinence (evolution 17, MCP-1, TNF-a, IL-10, and VEGF. These molecules were time of the EMU to baseline, age at baseline, and sex). The results chosen based on published results of previous studies regarding have been described with odds ratios (OR), 95% confidence both local and systemic biomarkers in uveitis and UME. Selected intervals (CI), and P-values. Linear relationships between immune mediators were determined by a Luminex platform defined biomarkers were assessed using Spearman tests. For all (Millipore’s MilliPlex Human Cytokine/Chemokine kit) used to tests, values of P <0.05 were considered statistically significant. measure cytokine and chemokine levels in serum samples using The R Studio statistical package (version 2.5) was used for an assay plate layout consisting of seven standards in duplicate statistical analysis. (3.2–2,000 pg/mL), one blank well (for background fluorescence subtraction), two internal quality control samples in duplicate, and 25-mL duplicates of each serum sample. The MilliPlex RESULTS method was performed as recommended by the manufacturer. Zero values were statistically handled as a third of the provided Sixty eyes of sixty patients were included. The mean age of the detection limit. group was 51.1 years (SD, 15.1; range, 21–89). The mean previous duration of UME at the inclusion time was 17.37 Treg Determination by Flow Cytometry months (SD, 31.5; range, 0–144 months). Table 1 shows FoxP3 levels in CD3+CD4+ cells from cryopreserved PBMCs were patients’ demographics and clinical data. The description of determined by flow cytometry. A live/dead Fixable Dead Cell Stain etiological diagnosis showed that systemic disorders were the (Invitrogen) was used to exclude dead cells from the analysis, predominant ones (n = 24), followed by isolated ocular followed by incubation for 30 min on ice with AF700-CD3 (clone syndromes (n = 21), and unclassified uveitis (n = 15). OKT3) and FITC-CD4 (OKT4) (eBiosciences). Cells were then The overall mean CST was 437.8 ± 122.2 (SD) mm at baseline washed, fixed and permeabilized before adding PE-FOXP3 (clone and decreased significantly after 1 month to 357.4 ± 118.9 mm 236A/E7) (eBiosciences) for 30 min. Appropriate isotype controls (P < 0.001), reaching 337.73 ± 135.5 mm at 6 months and 329.6 ± were included in each experiment. Flow cytometric analyses were 108.1 mm at 12 months (P < 0.001). Reduction in CST followed a performed on a Fortessa flow cytometer (BD Biosciences) with a similar pattern in systemically treated patients and non- Frontiers in Immunology | www.frontiersin.org 3 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME TABLE 1 | Baseline ophthalmic characteristics in eyes with uveitic macular edema. Baseline characteristics Eyes/Patients (n/n) 60/60 Evolution time to baseline (months, mean ± SD) 17 ± 31 Follow-up (months) 12 Age (years, mean ± SD) 51 ± 15 Female, n (%) 35 (58.3) Bilateral, n (%) 23 (38.3) Patterns of UME1 Cystoid, n (%) 47 (78.3) Diffuse, n (%) 14 (23.3) Subretinal fluid, n (%) 21 (35) Tractional component2, n (%) 18 (30) Anatomical classification (SUN) Anterior, n (%) 16 (26.6) Intermediate, n (%) 9 (15) Posterior, n (%) 22 (36.7) Panuveitis, n (%) 13 (21.7) Etiologic classification Unclassifiable (undifferentiated), n (%) 15 (25) Seronegative spondyloarthropathies, n (%) 12 (20) Ankylosing spondylitis, n 6 HLA-B27 + (without SA), n 3 Inflammatory bowel disease, n 1 Psoriatic arthritis, n 1 Reactive arthritis, n 1 White dot syndromes,n (%) 14 (23.3) Birdshot, n 13 Serpiginous, n 1 Other etiologies, n (%) 19 (31.6) Sarcoidosis, n 6 Vogt-Koyanagi-Harada, n 5 Juvenile idiopathic arthritis 2 Multiple sclerosis, n 2 Pars planitis, n 2 Sympathetic ophthalmia, n 1 IRVAN, n 1 Treatment/s for UME1 Topical, n (%) 2 (3.3) Local (only), n (%) 14(23.3) TCA (peri/intraocular), n 5 Dexamethasone implant, n 11 Systemic ± local, n (%) 40 (66.7) Oral Corticosteroids, n 37 Classic Immunomodulators, n 19 Biologics, n 13 Previous vitrectomy, n (%) 2 (3.3) 1 One eye may belong to more than one category (four eyes did not receive any treatment). 2It refers to qualitatively thickened vitreo-macular interface. UME, uveitis macular edema; SUN, standardization of uveitis nomenclature classification; SA, spondyloarthropaties; IRVAN, idiopathic retinal vasculitis, aneurisms and neuroretinitis syndrome; TCA, triamcinolone acetonide injections. systemically treated eyes, reaching a significant reduction already In the univariate analysis, qualitatively assessed significant at month 1 and maintaining a sustained reduction throughout tractional component of the UME was associated to non- the 12-month follow-up (Figure 1). sustained anatomical improvement, whereas presence of Cytokines and Tregs levels were compared between patients subretinal fluid was associated to sustained improvement in with and without systemic treatment. In our cohort, 66.7% of UME (Table 3). To evaluate prognostic factors for UME patients had received systemic treatment for their underlying outcome, variables were analyzed according to the pre-defined disease (non-infectious uveitis), and 33.3% had not received CST improvement event with a Step Backwise Logistic systemic treatment for at least 6 months before inclusion. Regression Model. Higher macular volume at baseline was the There were no significant differences in the levels of cytokines only clinical variable that precluded favorable UME outcome and Tregs between both groups (Table 2). However, differences (OR: 1.41, 95% CI: 0.97–2.03, P = 0.032). Interestingly, higher among specific treatments could not properly be evaluated due to baseline levels of IL-6 were associated with no sustained the reduced sample size in each specific treatment subgroup. anatomical improvement of UME at 12 months of follow-up Frontiers in Immunology | www.frontiersin.org 4 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME A B FIGURE 1 | Evolution of central subfield thickness (CST) in systemically treated eyes (A) and non-systemically treated eyes (B). Statistical analysis was conducted using the Wilcoxon test (*P < 0.05, **P < 0.001, ***P < 0.001 ****P < 0.0001, vs. baseline). TABLE 2 | Inflammatory mediators in patients with or without systemic treatment at baseline. Mediator, mean ± SD No systemic treatment Systemic treatment P-value n (eyes/patients) 20/20 40/40 TNF-a (pg/mL) 8.83 ± 6.17 9.39 ± 10.44 0.551 IL-1b (pg/mL) 0.91 ± 1.69 0.46 ± 0.52 0.432 IL-6 (pg/mL) 3.10 ± 5.86 6.44 ± 16.43 0.545 IL-8 (pg/mL) 13.66 ± 21.02 12.70 ± 11.57 0.960 IL-10 (pg/mL) 2.01 ± 6.52 0.52 ± 0.97 0.276 IL-17 (pg/mL) 10.7 ± 19.91 15.48 ± 30.68 0.778 MCP-1 (pg/mL) 617.41 ± 404.47 676.45 ± 409.43 0.846 VEGF (pg/mL) 275.58 ± 216.92 226.82 ± 155.29 0.389 Tregs (% CD3+ CD4+ PoxP3+) 3.78 ± 2.46 4.53 ± 2.76 0.392 TNF, tumor necrosis factor; IL, interleukin; MCP, monocyte chemoattractant protein; VEGF, vascular endothelial growth factor; Tregs, regulatory T lymphocytes. (OR: 0.67, 95% CI: 0.45–1.00); P = 0.042). In contrast, higher this is the first study that describes the association of Tregs with baseline levels of Tregs were associated with UME sustained anatomical outcome in UME. Previous studies have compared anatomical improvement (OR: 1.25, 95% CI: 1.12–2.56, P = Tregs levels between active uveitis, inactive uveitis and healthy 0.049). The final model showed good goodness-of-fit (Lemeshow control subjects. Some studies reported decreased levels of Tregs P = 0.13) (Figure 2). Given the observed associations, the inverse in active uveitis compared to inactive or quiescent status (9, 17) relationship between Tregs and IL-6 was evaluated. Spearman or healthy subjects (18), while others observed no differences correlation showed no linear association between baseline levels (19) or even increased levels. Indeed, a previous study by our of Tregs and IL-6 (r = 0.254, P = 0.109). group showed no differences in Tregs levels between patients with active non-infectious uveitis and healthy subjects (20). Regarding the potential association of UME and Tregs levels, DISCUSSION Yeh et al. did not observe differences between uveitis patients with and without UME (9). These discrepancies between studies In this prospective study, we evaluated the association of several could partly be attributed to the heterogeneity of uveitis subsets, inflammatory cytokines and Tregs with anatomical clinical immunophenotyping and to the different strategies used improvement of UME. We analyzed the systemic levels of to characterize Tregs. FoxP3 is the canonical transcription factor Tregs and eight serum immune mediators in a large group of for naturally-occurring Tregs and is enriched in human patients who presented with macular edema secondary to non- CD4+CD25hi T cells. Nevertheless, the CD4+CD25hi population infectious uveitis. Our findings show that high baseline serum does not necessarily capture all FoxP3+ cells, therefore we levels of IL-6 were associated with poorer anatomical response in defined CD3+CD4+FoxP3+ as Tregs. The existence of several UME, whereas high levels of circulating Tregs were associated subtypes of Tregs involved in the modulation of the with UME sustained resolution throughout 12-month follow up. inflammatory response, could also explain such discrepancies. The involvement of Tregs in controlling inflammation has Thymus originated tTregs inhibit T effector cell trafficking, gained considerable attention in uveitis as well as in other whereas inducible iTregs primarily prevent T cell priming by systemic autoimmune conditions (15, 16). To our knowledge, acting on antigen-presenting dendritic cells (21). The expression Frontiers in Immunology | www.frontiersin.org 5 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME TABLE 3 | Clinical variables in UME with and without anatomical improvement. Clinical variables at baseline Total No improvement Improvement1 P-value2 Total (eyes/patients) 60/60 38/38 22/22 Clinical imputs Age (years, mean ± SD) 51.1 ± 15.1 54 ± 14.7 46.1 ± 14.8 0.063 Female, n (%) 35 (58.3) 24 (63.2) 11 (50) 0.319 Evolution time (months, mean ± SD) 17.3 ± 31.5 12.5 ± 22.49 25.7 ± 42.2 0.357 Bilateral, n (%) 23 (38.3) 14 (36.8) 9 (40.9) 0.755 ACC (SUN) > 0.5+, n (%) 29 (48.3) 17 (44.7) 12 (54.5) 0.464 Keratic precipitates, n (%) Fines 10 (16.9) 7 (18.9) 3 (13.6) 0.725 Granulomatous 6 (10.2) 3 (8.1) 3 (13.6) Vitreous haze (NEI) > 0.5+, n (%) 29 (48.3) 20 (52.6) 9 (40.9) 0.381 Chorioretinal lesions (any), (%) 15 (25) 12 (31.6) 3 (13.6) 0.122 Anterior uveitis (SUN), n (%) 16 (26.7) 9 (23.7) 7 (31.8) 0.492 Unclassifiable uveitis, n (%) 7 (20.6) 5 (23.8) 2 (15.4) 0.654 Ocular comorbidities, n (%) Cataract 16 (26.7) 8 (21.1) 8 (36.4) 0.196 Glaucoma 11 (18.3) 7 (18.4) 4 (18.2) 0.982 Treatments for UME, n (%) None 4 (6.9) 4 (10.8) 0 (0) Local only 14 (23.3) 10 (27) 4 (19) 0.191 Systemic 40 (66.7) 23 (62.2) 17 (81) SD-OCT imputs Macular volume (mm3, mean ± SD) 11.9 ± 1.9 11.5 ± 1.3 12.6 ± 2.6 0.063 Thickened vitreo-macular interface, n (%) 18(30) 14 (36.8) 4 (18.2) 0.129 Unaltered ellipsoid layer, n (%) 35 (58.3) 25 (65.8) 10 (45.5) 0.124 Significant Tractional component, n (%) 7 (11.7) 7 (18.4) 0 (0) 0.032 UME cystoid component, n (%) 47 (78.3) 29 (76.3) 18 (81.8) 0.618 Diameter of major cyst (µm, mean ± SD) 330.1 ± 475.5 266.8 ± 208.8 433.5 ± 723.2 0.706 UME diffuse component, n (%) 14 (23.3) 11 (28.9) 3 (13.6) 0.177 SRF transverse extent (µm, mean ± SD) 363.1 ± 520.2 169 ± 271.6 621.9 ± 665.8 0.021 Choroidal thickness (µm, mean ± SD) 272.8 ± 97.7 258.5 ± 80.6 297.4 ± 119.9 0.137 Immune imputs Tregs (% CD3+ CD4+ PoxP3+) 4.2 ± 2.6 4.0 ± 2.3 4.9 ± 2.9 0.153 TNFa (pg/mL) 8.9 ± 8.2 10.1 ± 9.3 7.1 ± 5.9 0.136 IL-1b (pg/mL) 0.7 ± 1.4 1.0 ± 1.7 0.3 ± 0.3 0.075 IL-6 (pg/mL) 4.7 ± 11.9 6.2 ± 14.2 2.3 ± 6.7 0.064 IL-8 (pg/mL) 12.9 ± 18.5 14.7 ± 21.8 10.0 ± 11.1 0.281 IL-10 (pg/mL) 1.5 ± 5.4 1.1 ± 2.4 2.1 ± 8.3 0.281 IL-17 (pg/mL) 12.7 ± 25.4 9.5 ± 14.3 17.9 ± 36.9 0.589 MCP-1 (pg/mL) 645.6 ± 396.3 656.3 ± 410.8 628.4 ± 381.0 0.822 VEGF (pg/mL) 263.8 ± 195.8 282.5 ± 178.8 233.6 ± 221.7 0.127 1 Improvement, defined as per decrease ≥20% or reach <300 µm in central subfield thickness at 6 months and maintained through month 12. 2 Chi-square or Fisher (categorical), Mann-Whitney (continuous). ACC, anterior chamber cells; SUN standardization of uveitis nomenclature; NEI, National Eye Institute scale; UME, uveitic macular edema; SD-OCT, spectral-domain optical coherence tomography; SRF, subretinal fluid; TNF, tumor necrosis factor; IL, interleukin; MCP, monocyte chemoattractant protein; VEGF, vascular endothelial growth factor; Tregs, regulatory T cells. of the transcription factor Helios is a useful marker for the Gilbert et al. recently showed that clinical remission of non- identification of stable tTregs. However, we did not include infectious uveitis is characterized by an upregulation of peripheral Helios in our flow cytometry panel to characterize Tregs Tregs polarized toward TIGIT and T-bet (24). TIGIT is a co- subtype. Previous studies determined Tregs levels in patients inhibitory molecule expressed by Tregs. TIGIT+ Tregs seem to with and without UME in a cross-sectional manner, whereas the selectively inhibit Th17 and Th1, but not Th2 responses (25). Since present study is prospective in nature, with the aim to investigate it is known that the Th17 and Th1 subsets are pivotal in the the predictive value of Tregs levels in the anatomical response pathogenesis of autoimmune disorders, TIGIT could be a after 12 months. biomarker of stable, functionally suppressive Tregs in such The differences observed in Tregs levels according to the disorders. Even though we did not specifically analyze the anatomical event of UME improvement suggest that Tregs could suppressive function of Tregs, we observed an independent be associated with the protective processes underlying UME association between increased Tregs and anatomical improvement resolution. In line with our findings, preclinical evidence shows or resolution of UME, which was a 25% more frequently found for that in experimental autoimmune uveoretinitis (EAU), the adoptive each unitary increase in Tregs log levels. Almost 67% of recruited transfer of Tregs appears to confer protection from uveitis induced patients received systemic treatment showing a good clinical ocular by the uveitogenic retinal antigen interphotoreceptor binding response, supporting the fact that systemic peripheral blood Tregs protein (IRBP) (22), and another study showed that retinal and their associated microenvironment influence ocular immunity antigen-specific Foxp3+ Tregs contribute to the natural resolution and that, therefore, peripheral blood biomarkers can be useful in of EAU and the maintenance of remission (23). Interestingly, predicting ocular outcomes. Frontiers in Immunology | www.frontiersin.org 6 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME FIGURE 2 | Uni and multivariate regression model of variables as predictors of anatomic outcome in uveitic macular edema. Outcome: “sustained anatomical improvement”, defined as per decrease ≥20% or reaches <300 µm in central subfoveal thickness at 6 months and maintained through month 12. All covariates (P < 0.20) included in the multivariable model are shown. Quantitative variables were transformed to log scale. Statistically significant associations (P-value < 0.05) are noted in bold. Hosmer-Lemeshow goodness-of-fit test, P = 0.13. Abbreviations: CI, confidence interval; OR, odds ratio; UME, uveitic macular edema; SRF, subretinal fluid; TNF, tumor necrosis factor; IL, interleukin; VEGF, vascular endothelial growth factor; Tregs, regulatory T lymphocytes. On the other hand, the cytokine milieu in non-infectious blood-retinal barrier in vitro (38), a key feature of the uveitis has been extensively investigated (26–30). In the present pathophysiology of macular edema. Our current study actually study our focus was investigating the relationship between serum reinforces the concept of targeting systemic IL-6 in UME, as cytokines and anatomical improvement of UME. We therefore patients with higher baseline serum levels of IL-6 showed a less selected defined outcomes that may influence UME prognosis favorable anatomical response. and can be easily and objectively assessed. BCVA, although a Despite the association of increased Tregs and reduced IL-6 major clinical end result of UME, is not one of these features, levels with UME anatomical improvement, no linear association since it can be easily biased by other parameters such as UME between IL-6 and Tregs was observed. The exact relationship duration, presence of cataract, retinal scarring, or macular non- between IL-6 and Tregs in modulating uveitis is not clearly perfusion. We investigated whether anatomical improvement of understood. IL-6 is involved in the differentiation of CD4+ T UME could be predicted by immunological variables besides cells into Th17 cells known to play a pivotal role in uveitis. Tregs after adjusting for covariates in the model. The Moreover, blockade of IL-6 signaling not only suppresses Th17 multivariate model showed that higher baseline levels of IL-6 but also promotes Tregs in EAU (39). In other autoimmune were associated with lower CST improvement, which was a 33% disorders such as psoriasis, IL6 signaling prevents immune less frequent for each unitary increase in IL-6 log-levels. IL-6 is a suppression by Tregs (40). Other authors have shown that IL6 cytokine produced by several immune cells, in response to overproduction in vivo inhibits inducible Treg generation from molecular patterns and affects multiple inflammatory cells and naïve T cells, but does not affect the development and function of pathways. IL-6 is responsible for the induction of acute-phase natural Tregs (41). The inverse association of IL-6 and Tregs proteins, differentiation of adaptive T cell responses, trafficking might occur only at the local site within the eye, but not in the of acute and chronic inflammatory cells, regulation of systemic circulation, which could explain why we did not observe homeostasis, and tissue regeneration (31). In the eye, significant any association. Alternatively, the limited number of samples elevation of IL-6 has been observed in ocular fluids (vitreous and could also explain the lack of association. Thus, how IL6 aqueous) derived from retinal vein occlusion, diabetic macular modulates Treg function in the eye warrants further investigation. edema, and refractory/chronic uveitis patients (32–34). Our group Regarding the therapeutic approach, one may expect a and others have previously reported that in patients with UME relationship between systemic treatment and serum cytokine refractory to conventional therapies, systemic inhibition of IL-6 levels, as immunomodulatory agents are used to treat non- signaling with the IL-6 receptor (IL-6R) monoclonal antibody infectious uveitis. However, we did not observe significant tocilizumab (TCZ–Actemra; Hoffmann-La Roche Ltd, Basel, differences in cytokine levels between none of the initial Switzerland) was beneficial at 6, 12, and 24 months (35, 36), in treatment modalities, either local, systemic, or a combination particular with regards to achieving a reduction in CST. Indeed, of both. IL-6 signaling blockade suppresses EAU (37), and it has been The main limitations of the present study include the sample recently described that IL-6 reversibly disrupts the integrity of the size of eyes included in the analysis, the heterogeneity of uveitic Frontiers in Immunology | www.frontiersin.org 7 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME conditions and immunomodulatory treatments, and the Clı́nic de Barcelona 2013/8574; Comité de É tica de la Investigació n interindividual variability of cytokine levels. Despite these con Medicamentos de Euskadi, Hospital Universitario Cruces concerns, the strengths of the study include a prospective PI201406; and Comité É tico de Investigació n Clı́nica del design, the use of standardized masked data collection Hospital Clı́nico San Carlos de Madrid 13/244-E. The patients/ protocols, a centralized center for the cytokine and Tregs participants provided their written informed consent to participate determination, and the recruitment from multiple uveitis in this study. referral centers. In conclusion, the results presented herein suggest that higher levels of Tregs may contribute to sustained anatomical improvement of UME, whereas increased systemic levels of IL- AUTHOR CONTRIBUTIONS 6 may be a prognostic factor of poor anatomical outcome of VL, MM, AF, DD-V, AA, and BM contributed to the design of UME. These findings could open the opportunity to more the study and experiments. JM, VL, AF, DD-V, CE, JA, BB, MS, efficient and personalized therapeutic approaches to improve AA, and BM performed the experiments, data capture, and long-term visual prognosis in patients with macular edema analysis. JM, VL, MF-R, and BM performed the data secondary to non-infectious uveitis. interpretation. All authors contributed to the article and approved the submitted version. DATA AVAILABILITY STATEMENT The raw data supporting the conclusions of this article will be FUNDING made available by the authors, without undue reservation. This work was supported by the Ministry of Science and Innovation of Spain, ‘Instituto de Salud Carlos III,’ ‘Fondo de Investigació n Sanitaria’ (PI13/00217, PI17/00316, and RD16/0008), ETHICS STATEMENT and funds FEDER “Una manera de hacer Europa.” We thank the support of the Generalitat of Catalunya (Secretaria d’Universitats The studies involving human participants were reviewed and i Recerca del Departament d’Economia i Coneixement de la approved by Comité É tico de Investigació n Clı́nica del Hospital Generalitat, 2017 SGR 0701. 11. Siepmann K, Biester S, Plšková J, Muckersie E, Duncan L, Forrester JV. REFERENCES CD4+CD25+ T regulatory cells induced by LPS-activated bone marrow 1. Nussenblatt RB. The natural history of uveitis. Int Ophthalmol (1990) 14(5- dendritic cells suppress experimental autoimmune uveoretinitis in vivo. 6):303–8. doi: 10.1007/BF00163549 Graefe’s Arch Clin Exp Ophthalmol (2007) 245(2):221–9. doi: 10.1007/ 2. Suttorp-Schulten MSA, Rothova A. The possible impact of uveitis in s00417-006-0356-9 blindness: A literature survey. Br J Ophthalmol (1996) 80(9):844–8. 12. Levin MH, Pistilli M, Daniel E, Gangaputra SS, Nussenblatt RB, Rosenbaum doi: 10.1136/bjo.80.9.844 JT, et al. Incidence of visual improvement in uveitis cases with visual 3. Dick AD. Doyne lecture 2016: Intraocular health and the many faces of impairment caused by macular edema. Ophthalmology (2014) 121:588– inflammation. Eye (2017) 31(1):87–96. doi: 10.1038/eye.2016.177 95.e1. doi: 10.1016/j.ophtha.2013.09.023 4. Munk MR, Kiss CG, Steiner I, Sulzbacher F, Roberts P, Kroh M, et al. 13. Ossewaarde-Van Norel A, Rothova A. Clinical review: Update on treatment of Systematic correlation of morphologic alterations and retinal function in eyes inflammatory macular edema. Ocul Immunol Inflammation (2011) 19(1):75– with uveitis-associated cystoid macular oedema during development, 83. doi: 10.3109/09273948.2010.509530 resolution and relapse. Br J Ophthalmol (2013) 97:1289–96. doi: 10.1136/ 14. Jabs DA, Nussenblatt RB, Rosenbaum JT, Atmaca LS, Becker MD, Brezin AP, bjophthalmol-2012-303052 et al. Standardization of uveitis nomenclature for reporting clinical data. 5. Rothova A. Medical treatment of cystoid macular edema. Ocul Immunol Results of the first international workshop. Am J Ophthalmol (2005) 140 Inflammation (2002) 10(4):239–46. doi: 10.1076/ocii.10.4.239.15589 (3):509–16. doi: 10.1016/j.ajo.2005.03.057 6. Jeon S, Lee WK, Jung Y. Changes in the intraocular cytokine levels after 15. Rizzo LV, Xu H, Chan CC, Wiggert B, Caspi RR. IL-10 has a protective role in intravitreal bevacizumab in uveitic macular edema. Ocul Immunol experimental autoimmune uveoretinitis. Int Immunol (1998) 10(6):807–14. Inflammation (2012) 20(5):360–4. doi: 10.3109/09273948.2012.709576 doi: 10.1093/intimm/10.6.807 7. Valencia X, Lipsky PE. CD4+CD25+FoxP3+regulatory T cells in autoimmune 16. Roncarolo MG, Bacchetta R, Bordignon C, Narula S, Levings MK. Type 1 T diseases. Nat Clin Pract Rheumatol (2007) 3(11):619–26. doi: 10.1038/ regulatory cells. Immunol Rev (2001) 182:68–79. doi: 10.1034/j.1600- ncprheum0624 065X.2001.1820105.x 8. Ruggieri S, Frassanito MA, Dammacco R, Guerriero S. T reg lymphocytes in 17. Chen L, Yang P, Zhou H, He H, Ren X, Chi W, et al. Diminished frequency autoimmune uveitis. Ocul Immunol Inflammation (2012) 20(4):255–61. and function of CD4+CD25high regulatory T cells associated with active doi: 10.3109/09273948.2012.681830 uveitis in Vogt-Koyanagi-Harada syndrome. Invest Ophthalmol Vis Sci (2008) 9. Yeh S, Li Z, Forooghian F, Hwang FS, Cunningham MA, Pantanelli S, et al. 49(8):3475–82. doi: 10.1167/iovs.08-1793 CD4 +foxp3 + T-regulatory cells in noninfectious uveitis. Arch Ophthalmol 18. Sugita S, Yamada Y, Kaneko S, Horie S, Mochizuki M. Induction of regulatory (2009) 127(4):407–13. doi: 10.1001/archophthalmol.2009.32 T cells by infliximab in behçet’s disease. Invest Ophthalmol Vis Sci (2011) 52 10. Keino H, Takeuchi M, Usui Y, Hattori T, Yamakawa N, Kezuka T, et al. (1):476–84. doi: 10.1167/iovs.10-5916 Supplementation of CD4+CD25+ regulatory T cells suppresses experimental 19. Commodaro AG, Peron JPS, Genre J, Arslanian C, Sanches L, Muccioli C, et al. autoimmune uveoretinitis. Br J Ophthalmol (2007) 91(1):105–10. IL-10 and TGF-b immunoregulatory cytokines rather than natural regulatory doi: 10.1136/bjo.2006.099192 T cells are associated with the resolution phase of Vogt-Koyanagi-Harada Frontiers in Immunology | www.frontiersin.org 8 September 2020 | Volume 11 | Article 579005 Matas et al. Tregs and IL-6 in UME (VKH) syndrome. Scand J Immunol (2010) 72(1):31–7. doi: 10.1111/j.1365- edema. Curr Diabetes Rep (2013) 13:476–80. doi: 10.1007/s11892-013- 3083.2010.02401.x 0382-z 20. Molins B, Mesquida M, Lee RWJ, Llorenç V, Pelegrı́n L, Adá n A. Regulatory T 33. Noma H, Funatsu H, Mimura T, Harino S, Hori S. Vitreous Levels of cell levels and cytokine production in active non-infectious uveitis: In-vitro Interleukin-6 and Vascular Endothelial Growth Factor in Macular Edema effects of pharmacological treatment. Clin Exp Immunol (2015) 179(3):529– with Central Retinal Vein Occlusion. Ophthalmology (2009) 116:87–93. 38. doi: 10.1111/cei.12479 doi: 10.1016/j.ophtha.2008.09.034 21. Shevach EM, Thornton AM. tTregs, pTregs, and iTregs: Similarities and 34. van Kooij B, Rothova A, Rijkers GT, de Groot-Mijnes JDF. Distinct cytokine and differences. Immunol Rev (2014) 259(1):88–102. doi: 10.1111/imr.12160 chemokine profiles in the aqueous of patients with uveitis and cystoid macular 22. Silver PB, Agarwal RK, Su S-B, Suffia I, Grajewski RS, Luger D, et al. edema. Am J Ophthalmol (2006) 142:192–4. doi: 10.1016/j.ajo.2006.02.052 Hydrodynamic Vaccination with DNA Encoding an Immunologically 35. Mesquida M, Molins B, Llorenç V, Herná ndez MV, Espinosa G, Sainz de la Privileged Retinal Antigen Protects from Autoimmunity through Induction Maza M, et al. Twenty-Four Month Follow-Up Of Tocilizumab Therapy For of Regulatory T Cells. J Immunol (2007) 179(8):5146–58. doi: 10.4049/ Refractory Uveitis-Related Macular Edema. Retina (2018) 38(7):1361–70. jimmunol.179.8.5146 doi: 10.1097/IAE.0000000000001690 23. Silver PB, Horai R, Chen J, Jittayasothorn Y, Chan C-C, Villasmil R, et al. 36. Mesquida M, Molins B, Llorenç V, Sainz de la Maza M, Adá n A. Long-term Retina-Specific T Regulatory Cells Bring About Resolution and Maintain effects of tocilizumab therapy for refractory uveitis-related macular edema. Remission of Autoimmune Uveitis. J Immunol (2015) 194(7):3011–9. Ophthalmology (2014) 121:2380–6. doi: 10.1016/j.ophtha.2014.06.050 doi: 10.4049/jimmunol.1501038 37. Rochfort KD, Collins LE, Murphy RP, Cummins PM. Downregulation of 24. Gilbert RM, Zhang X, Sampson RD, Ehrenstein MR, Nguyen DX, Chaudhry M, blood-brain barrier phenotype by proinflammatory cytokines involves et al. Clinical remission of sight-threatening non-infectious uveitis is NADPH oxidase-dependent ROS generation: Consequences for characterized by an upregulation of peripheral T-regulatory cell polarized interendothelial adherens and tight junctions. PLoS One (2014) 9(7): towards T-bet and TIGIT. Front Immunol (2018) 9:907. doi: 10.3389/ e101815. doi: 10.1371/journal.pone.0101815 fimmu.2018.00907 38. Mesquida M, Drawnel F, Lait PJ, Copland DA, Stimpson ML, Llorenç V, et al. 25. Joller N, Lozano E, Burkett PR, Patel B, Xiao S, Zhu C, et al. Treg cells Modelling macular edema: The effect of IL-6 and IL-6r blockade on human expressing the coinhibitory molecule TIGIT selectively inhibit blood-retinal barrier integrity in vitro. Transl Vis Sci Technol (2019) 8(5):32. proinflammatory Th1 and Th17 cell responses. Immunity (2014) 40(4):569– doi: 10.1167/tvst.8.5.32 81. doi: 10.1016/j.immuni.2014.02.012 39. Hohki S, Ohguro N, Haruta H, Nakai K, Terabe F, Serada S, et al. Blockade of 26. Torun N, Callizo J, Orlic N, Scherer M, Hartmann C, Pleyer U. Serum interleukin-6 signaling suppresses experimental autoimmune uveoretinitis by cytokine receptor levels in noninfectious uveitis. Ophthalmic Res (2005) 37 the inhibition of inflammatory Th17 responses. Exp Eye Res (2010) 91:162–70. (2):112–6. doi: 10.1159/000084271 doi: 10.1016/j.exer.2010.04.009 27. Takase H, Futagami Y, Yoshida T, Kamoi K, Sugita S, Imai Y, et al. Cytokine 40. Goodman WA, Levine AD, Massari JV, Sugiyama H, McCormick TS, Cooper profile in aqueous humor and sera of patients with infectious or noninfectious KD. IL-6 Signaling in Psoriasis Prevents Immune Suppression by Regulatory uveitis. Invest Ophthalmol Vis Sci (2006) 47(4):1557–61. doi: 10.1167/iovs.05- T Cells. J Immunol (2009) 183(5):3170–6. doi: 10.4049/jimmunol.0803721 0836 41. Fujimoto M, Nakano M, Terabe F, Kawahata H, Ohkawara T, Han Y, et al. 28. Mesquida M, Molins B, Llorenç V, Sainz De La Maza M, Hernandez MV, The Influence of Excessive IL-6 Production In Vivo on the Development and Espinosa G, et al. Proinflammatory cytokines and c-reactive protein in uveitis Function of Foxp3 + Regulatory T Cells. J Immunol (2011) 186(1):32–40. associated with behçet’s disease. Mediators Inflamm (2014) 2014:396204. doi: 10.4049/jimmunol.0903314 doi: 10.1155/2014/396204 29. Sainz-de-la-Maza M, Molins B, Mesquida M, Llorenç V, Zarranz-Ventura J, Conflict of Interest: The authors declare that the research was conducted in the Sala-Puigdollers A, et al. Interleukin-22 serum levels are elevated in active absence of any commercial or financial relationships that could be construed as a scleritis. Acta Ophthalmol (2016) 94(6):e395–9. doi: 10.1111/aos.13005 potential conflict of interest. 30. Molins B, Mesquida M, Llorenç V, Sainz de la Maza M, Adá n A. Elevated Serum Immune Mediators and Subclinical Inflammation in HLA-A29-associated Copyright © 2020 Matas, Llorenç, Fonollosa, Dı ́az-Valle, Esquinas, de la Maza, Birdshot Chorioretinopathy. Ocul Immunol Inflamm (2016) 24(6):647–652. Figueras-Roca, Artaraz, Berasategui, Mesquida, Adán and Molins. This is an doi: 10.3109/09273948.2015.1057601 open-access article distributed under the terms of the Creative Commons 31. Mesquida M, Molins B, Llorenç V, de la Maza MS, Adá n A. Targeting Attribution License (CC BY). The use, distribution or reproduction in other forums interleukin-6 in autoimmune uveitis. Autoimmun Rev (2017) 16(10):1079–89. is permitted, provided the original author(s) and the copyright owner(s) are credited doi: 10.1016/j.autrev.2017.08.002 and that the original publication in this journal is cited, in accordance with accepted 32. Owen LA, Hartnett ME. Soluble mediators of diabetic macular edema: the academic practice. No use, distribution or reproduction is permitted which does not diagnostic role of aqueous VEGF and cytokine levels in diabetic macular comply with these terms. Frontiers in Immunology | www.frontiersin.org 9 September 2020 | Volume 11 | Article 579005