Interleukin-4 Plasma Levels Stratified by Sex in Intermediate Age-Related Macular Degeneration and Geographic Atrophy

Abstract

Purpose: Chronic local inflammation underlies the pathogenesis of age-related macular degeneration (AMD) causing damage to the neurosensory retina. However, there is minimal research on systemic cell-mediated inflammation in AMD. Interleukin-4 (IL-4) is an immunoregulatory cytokine with an important role in modulating inflammation in chronic immune mediated disease. The purpose of this study was to: (1) investigate the role of systemic IL-4 in patients with intermediate AMD (iAMD) and in geographic atrophy (GA), an advanced form of AMD, compared to controls without AMD, and (2) determine if IL-4 levels are moderated by sex.

Methods: We examined plasma levels of IL-4 in patients with iAMD, GA, and controls without AMD included in the University of Colorado AMD registry (August 2014 to June 2021). Cases and controls were defined by multimodal imaging. IL-4 was measured by multiplex immunoassay. Data were analyzed using a nonparametric rank based linear regression model fit to IL-4.

Results: There were 199 patients with iAMD, 97 patients with GA, and 139 controls, with a percentage of female patients 61%, 55%, and 66%, respectively. We demonstrated significantly higher median IL-4 levels in GA (35.3; interquartile range [IQR] = 22.8–50.5) compared to iAMD (6.1; IQR = 2.2–11.3, P < 0.01) and controls (10.7; IQR = 5.0–16.8, P < 0.01). There were no significant differences in levels of IL-4 for cases and controls when stratified by sex.

Conclusions: These findings demonstrate a systemic immunological difference between iAMD and GA, indicating IL-4 may be a systemic biomarker for GA development.

Translational Relevance: The plasma biomarker IL-4 is significantly elevated in patients with GA.

Introduction

Age-related macular degeneration (AMD) is the leading cause of vision loss in individuals over the age of 50 years worldwide.^1–3 The natural history of AMD is characterized by progression from the early to intermediate stages, followed by late stage AMD.^1,3–5 Early (eAMD) and intermediate AMD (iAMD) are defined by drusen deposition between the basal retinal pigment epithelium (RPE) and Bruch's membrane (BM) with the development of RPE pigmentary abnormalities in intermediate AMD.^5,6 The two forms of advanced AMD, neovascular AMD (nvAMD) and geographic atrophy (GA), are responsible for the majority of AMD-related visual damage.^1–3,6,7 

Patients with intermediate-stage AMD have up to a 47% 5-year risk of progression to the advanced stages of AMD.^8,9 Advanced AMD affects 1% to 3% of the total global population,¹ with GA accounting for approximately 90% of all cases.¹⁰ As described by other authors,^1–3,6,7 the hallmark histological finding in GA is the presence of atrophic lesions in the outer retina caused by the loss of photoreceptors, RPE, and choriocapillaris. This progressive destruction of neurosensory tissue eventually leads to irreversible vision loss.^1–3,6,7 

There is a consensus that dysregulation of local^1,3,10 and systemic^11–13 inflammatory pathways contribute to the pathogenesis of AMD. However, the role of systemic inflammation in AMD is understudied and is a focus of research from our group.^14–16 The overarching hypothesis for our research is that systemic inflammation contributes to local ocular events in AMD and identification of systemic biomarkers may lead to earlier diagnosis and targeted therapies to prevent further visual damage. In research to date, we have demonstrated a significant role for systemic inflammatory markers, specifically cytokines,^14,16 chemokines,^17–19 and complement factors,^20–22 with sex-specific modulation of the complement system²² and RANTES^18,19 in AMD. 

IL-4 is a pleotropic immunomodulatory cytokine with an important role in regulating the pathogenic immune response in chronic inflammatory disease.^23–28 To our knowledge, systemic IL-4 characterization data in iAMD²⁹ and GA is still emerging. Thus, the purpose of the research described herein was to investigate the role of IL-4 specifically in iAMD and GA to further define inflammatory pathways and potential systemic biomarkers involved in disease pathophysiology. The objectives of this study were to: (1) determine plasma levels of IL-4 in patients with iAMD and GA compared to controls without AMD, and (2) evaluate if concentrations of IL-4 were moderated by sex. 

Methods

Participants and Study Design

This study was conducted using samples and records from patients with iAMD, GA, and controls without AMD who are part of the University of Colorado AMD research registry. The registry, developed in the Department of Ophthalmology at the University of Colorado,^21,30,31 is approved by the Colorado Multiple Institutional Review Board and conforms with the Declaration of Helsinki. Patients with AMD who attend the retina clinic at the UCHealth Sue Anschutz-Rogers Eye center are invited to participate in the registry. All patients who agreed to participate in the registry provide written informed consent for collection of plasma and serum for biomarker studies, review of pertinent medical history, and examination of multimodal imaging (color fundus photography, autofluorescence, and ophthalmic coherence tomography). Recruitment into this longitudinal cohort is ongoing. 

Inclusion and Exclusion Criteria

The inclusion and exclusion criteria and informed consent are described in detail elsewhere.^18,19,22 In summary, inclusion criteria for case subjects included in the registry are aged between 55 and 99 years, unilateral or bilateral AMD, and capacity for informed consent. Exclusion criteria include: branch and central retinal vein occlusion with severe macular damage, pan-retinal photocoagulation or anti-VEGF injections for diabetic retinopathy, active ocular inflammatory disease, treatment with select immunosuppressives, serious mental health or terminal illnesses, and a decrease in visual acuity secondary to a severe pre-existing retinal disease except AMD. Control patients are patients who previously underwent cataract surgery at least 1 month prior to enrollment with no evidence of retinal disease confirmed by multimodal imaging. All data for the registry are entered into REDCap, a secure web-based institutional review board (IRB)-approved database. 

Image Review

For the purpose of this study, we concentrated on patients with GA, iAMD, and controls included in the registry between August 2014 and June 2021. Multimodal image review was conducted by two vitreoretinal specialists for cases and controls (authors N.M. and M.M.) with a third vitreoretinal specialist (author T.D.C.F.) to resolve discrepancies. Images were classified using the classification described by the Beckman Initiative for Macular Research Classification Committee.³² GA was further characterized by the Classification of Atrophy Meeting (CAM) criteria demonstrating complete RPE and outer retinal atrophy.^33,34 The following criteria were used to define complete RPE and outer retinal atrophy: (a) a region of choroidal hyper transmission greater than or equal to 250 µm in diameter; (b) an area of attenuation or disruption to the RPE of greater than or equal to 250 µm in diameter; and (c) evidence of overlying photoreceptor degeneration, all in the absence of evidence of a tear in the RPE.^33,34 

Blood Sample Collection and Processing

Ethylenediaminetetraacetic acid (EDTA) plasma samples were spun at 3000 rpm for 10 minutes in a chilled (4°C) centrifuge. The plasma was pipetted into aliquots and immediately stored in a −80°C freezer. The plasma aliquots were transferred to the laboratory for biomarker measurement. 

The R&D Systems Human Magnetic Luminex Performance High Sensitivity Assay Base Kit was used. The manufacturer of this approved assay performed the aspects of validation necessary to meet the requirements of regulatory compliance. Upon receipt of this manufacturer’s validated assay, all aspects of verification outlined by the College of American Pathologists (CAP) were performed to confirm assay performance and reproducibility, as described by other authors.^35–37 Manufacturer’s standard and control materials were used to confirm precision and accuracy of the assay and the possibility of sample matrix interference was analyzed. The linearity of the assay range was assessed and lower and upper limits of detection were evaluated. The clinical reportable range (CRR) was derived from the lower limit of detection (LLOD) and upper limit of detection (ULOD). The CRR was set at: 5.6 pg/mL and the LLOD for the IL-4 high sensitivity magnetic Luminex performance assay was set at: 2.8 pg/mL. Patient dilution linearity was assessed and used to determine how much a sample can be diluted and still generate reliable results. Sample dilution was used in the event a sample result was too high and fell outside of the linear range of the assay's reportable range. 

Measurement of Plasma Biomarkers

Levels of the plasma biomarker IL-4 were measured at the Clinical Translational Research Core laboratory at the Children's Hospital Colorado. Magnetic bead-based multiplex assays produced by R&D Systems with analyte-specific antibodies were analyzed on the Luminex FlexMap, a dual-laser suspension array platform. All samples were analyzed in duplicate with a threshold coefficient of variance of less than 15%. 

Statistical Analysis

Patient characteristics were compared between groups using an analysis of variance (ANOVA) for continuous variables and the chi-squared test or Fisher's exact test for categorical variables. Levels of IL-4 were fit with a regression model using a rank transformation that included an AMD classification by sex interaction, age, and assay run date as covariates. Least square means were used to analyze pairwise comparisons for all sexes by study group comparisons. All analyses were conducted with SAS version 9.4 (The SAS Institute, Cary, NC, USA). 

Results

In Table 1, we show select demographic characteristics and comorbidities across our three study groups: iAMD cases, GA cases, and controls. The control cohort had the highest proportion of female participants (66%), followed by the iAMD (61%) and GA cohorts (55%), although there were no significant differences across groups. The mean age of patients was highest in the GA group (82.3, SD = 7.0), followed by iAMD (76.4, SD = 6.9), and controls (74.2, SD = 4.7, P < 0.01 overall ANOVA comparison). Additionally, there was a greater frequency of family history of AMD reported in patients with iAMD, followed by patients with GA and controls (P < 0.01). Patients with GA also had a significantly higher prevalence of treated hypertension (73%, P < 0.01) in comparison to controls (54%) and patients with iAMD (53%). 

Table 1.

 

Differences in Clinical Characteristics Across AMD Groups

 

In Table 2 and the Figure, we demonstrate systemic levels of IL-4 for cases of iAMD and GA and controls. IL-4 was significantly higher in the GA cohort (median = 35.3, interquartile range [IQR] = 22.8–50.5) compared to the iAMD group (median = 6.1, IQR = 2.2–11.3) and controls (median = 10.7, IQR = 5.0–16.8, P < 0.01). As shown in Table 2 and the Figure, male subjects had higher levels of IL-4 (median = 39.0, IQR = 24.8–53.4) compared to female subjects (median = 34.9, IQR = 20.5–49.4) in the GA cohort when stratified by sex, P = 0.10, although this difference was not significant. There were no significant differences in levels of IL-4 between male and female subjects among the iAMD, GA, and control groups. Results were also adjusted for the possible confounding variables of age and batch effects. Plasma levels and P values for IL-4 were unchanged after a regression analysis with age and assay run date as covariates. There was no association between IL-4 levels and treated hypertension or family history of AMD, however, a sensitivity analysis including these factors as covariates in the regression model did not change the results. 

Table 2.

 

IL-4 Levels in iAMD, GA, and Controls With a Sex by AMD Group Interaction

 

Figure.

 

Plasma levels of IL-4 with a sex by AMD group interaction. Boxplot showing IL-4 levels for iAMD, GA, and control patients and IL-4 levels across groups when stratified by sex. The box extends to the 25th and 75th percentiles, the line represents the median value. Individual colored circles/plus signs illustrate raw values. Female patients are represented as blue circles and male patients are represented as red plus signs.

Discussion

In this study, we demonstrate significantly elevated levels of IL-4 in GA compared to the iAMD and control groups. Moreover, we found slightly higher levels of IL-4 in male subjects compared to female subjects among the GA cases, however, this difference did not reach statistical significance likely due to the small sample size. 

IL-4 is a member of the Th2 cytokine family and is secreted from Th2 cells, basophils, mast cells, and type II innate lymphoid cells. IL-4 has a central role in promoting Th2 cell differentiation³⁸ and antagonizes Th1 mediated inflammation by inhibiting secretion of pro-inflammatory cytokines and inducing macrophage polarization to an M2 anti-inflammatory phenotype.²³ Systemic IL-4 is elevated in a number of non-ocular pathologies, including systemic lupus erythematosus (SLE),^39,40 active psoriasis vulgaris,⁴¹ tropical endomyocardial fibrosis,⁴² atopic dermatitis, and atopic rhinitis⁴³ with a positive association between levels of IL-4 and disease severity in multiple sclerosis⁴⁴ and asthma-chronic obstructive pulmonary disease (COPD) overlap syndrome.⁴⁵ In ocular pathologies, a recent study by Gotfredsen et al.,²⁹ compared serum levels of IL-4 between patients with myeloproliferative disorders with and without drusen, iAMD, and nvAMD. These authors did not find any statistically significant differences in iAMD compared to the other groups. Our findings are in contrast to these results which may be due to the larger sample sizes in our study and different ethnic background of our study population. 

Our study showed an increased concentration of plasma IL-4 in GA compared to iAMD and control subjects. One possible explanation for our findings is an attempt to downregulate the systemic immune response as the disease progresses from iAMD to GA. It is unclear whether the specific driver of IL-4 upregulation is caused by disease pathogenesis or is a byproduct of unrelated systemic factors. We suggest a potential pathway underlying this finding is polarization of macrophages to the M2 phenotype. In ocular pathology, accumulation of byproducts from retinal damage causes a local inflammatory response which results in the recruitment of peripheral blood monocytes to the retina through permeable choroidal vessels.^46,47 We propose a similar mechanism to recruit systemic IL-4 to the retina, thus inducing a shift to the M2 phenotype. Prior studies have demonstrated M1 macrophage polarization earlier in disease pathogenesis promotes the development of GA.⁴⁸ However, M2 macrophages are thought to have a protective function in the choroid and retina, therefore, a shift to the M2 phenotype may represent an attempt to clear damaged cells and inhibit local inflammation as GA progresses.^49,50 Given the role of M2 macrophages in promoting retinal neovascularization,^51,52 our findings must be reconciled with concentrations of IL-4 in nvAMD to further speculate on the pathophysiologic mechanisms. 

It is also possible IL-4 may play a pathogenic role in the progression of AMD given its role in mast cell proliferation and degranulation.⁵³ Bhutto et al.⁵⁴ demonstrated higher choroidal mast cells (MCs) in donor eyes with early AMD (eAMD) and GA compared to aged controls. MCs were more frequently distributed in the choriocapillaris in GA compared to the choroidal stroma in aged controls, with increased loss of the choriocapillaris in areas with the greatest degranulated MCs.⁵⁴ GA eyes also showed increased tryptase staining in MCs located in areas of RPE atrophy and atrophic borders near the choriocapillaris.⁵⁵ In contrast, tryptase in aged controls was largely found in MCs in Sattler's layer of the choroidal stroma.⁵⁵ Tryptase in GA eyes was also distributed diffusely around MCs in the stroma, implying MC degranulation in GA.⁵⁵ Tryptase has been implicated in localized tissue destruction in conditions of chronic inflammation in the intestines^56,57 and in atherosclerotic arteries.^58,59 One mechanism which might explain our finding of elevated plasma IL-4 in GA is the diffusion of systemic IL-4 to the choroid and retina, mediating MC production, and degranulation. 

We found higher levels of IL-4 in male subjects compared to female subjects among the GA cases, although the difference was not statistically significant. Sex specific differences in epidemiology, pathophysiology, and disease progression are well established for autoimmune⁶⁰ and ocular pathologies, including AMD.^43,61,62 The underlying etiology is likely complex and multifactorial, encompassing epigenetic and chromosomal linked mechanisms⁶³ and differences in cellular and molecular function.⁶⁴ The possible sex specific differences in GA cases in our study may result from the immunomodulatory effect of testosterone in male subjects. As comprehensively reviewed by Mohamed et al.,⁶⁵ endogenous testosterone levels in male subjects were negatively correlated with concentrations of soluble inflammatory cytokines. Exogenous testosterone therapy may also induce a shift in the systemic cytokine balance by suppressing plasma expression of pro-inflammatory cytokines and enhancing expression of anti-inflammatory cytokines.^65,66 The potential sex difference in IL-4 concentrations in this study suggests the importance of sex as a biologic modifier in chronic inflammatory disease. 

Limitations of this study include cross-sectional measurement of cytokines at a single time point during registry enrollment. Our cohort is also recruited from a retinal service and may not be representative of the inflammatory milieu of all patients with iAMD or GA. Additionally, we do not have sufficient data on other plasma cytokines in this cohort which may provide further insight on disease mechanisms. Furthermore, we did not examine risk factors, such as hormonal profiles, which have sex-dependent effects on disease incidence and immunological function.⁶⁷ 

Our results demonstrate significantly increased plasma IL-4 in GA compared to iAMD, reflecting a marker of immunologic difference between GA and iAMD. Further research is required to evaluate this systemic pathway to better understand disease pathophysiology and define potential biomarkers and targeted therapies for this vision disabling disease.  

Acknowledgments

Supported by the Greenwald Family Research Fund, a Research to Prevent Blindness grant to the Department of Ophthalmology, University of Colorado, the Frederic C. Hamilton Macular Degeneration Center, Sue Anschutz-Rogers Eye Center Research Fund, NIH/NCATS Colorado CTSA Grant Number UL1 TR002535, and in part by the National Eye Institute of the National Institutes of Health under award number R01EY032456 (A.M.L.) 

The University of Colorado Retina Research Group: Melanie Akau, OD, Rebecca Baldermann, BA, Karen L. Christopher, MD, Richard Davidson, MD, Talisa E. De Carlo Forest, MD, C. Rob Graef, OD, Ramya Gnanaraj, MD, Anne M. Lynch, MB, BCH, BAO, MSPH, Naresh Mandava, MD, Niranjan Manoharan, MD, Marc T. Mathias, MD, Scott N. Oliver, MD, Jeffery L. Olson, MD, Alan G. Palestine, MD, Jennifer L. Patnaik, PhD, MHS, Jesse M. Smith, MD, and Brandie D. Wagner, PhD. 

Disclosure: V. Rajeswaren, None; B.D. Wagner, None; J.L. Patnaik, None; N. Mandava, None; M.T. Mathias, None; N. Manoharan, None; T.E. De Carlo Forest, None; R. Gnanaraj, None; A.M. Lynch, None; A.G. Palestine, None