92-Plex Inflammatory Proteomic Signatures in Aqueous Humor of High Myopia and Candidate Biomarkers Authors: Ming Yang, Weizhen Wu, Ning Dong, Min Li Published: 2026 ABSTRACT ObjectiveIn this exploratory study with a limited sample size, we aimed to investigate inflammation-related proteins in the aqueous humor of high myopia (HM) patients, analyze their correlations with ocular parameters, and screen for potential biomarkers.Methods20 age-related cataract patients undergoing surgery (Jan 1, 2025 – Feb 1, 2026) were enrolled: HM group (n=12) and non-HM control (n=8). Aqueous humor was collected preoperatively; 92 inflammation-related proteins were quantified using Olink Target 96 Inflammation Panel. Differential expression analysis, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis, protein-protein interaction (PPI) network analysis, and Spearman correlation analysis were performed to explore the biological functions of differentially expressed proteins and their relationships with ocular parameters.ResultsNine differentially expressed proteins (DEPs) (FDR q<0.05, |log2FC|>0.585) were upregulated in HM vs control, most prominently HGF, CST5, CCL19. Enrichment analyses revealed that the DEPs were mainly enriched in the cytokine-cytokine receptor interaction pathway, chemokine signaling pathway, and immune-inflammatory response-related terms. PPI network analysis identified TWEAK, CXCL5, and CCL19 as core proteins. Exploratory correlation analysis identified several associations of nominal significance (uncorrected p < 0.05) between cytokines and ocular parameters in both HM and non-HM groups (eg, DNER and TWEAK with central corneal thickness, CXCL5 with macular center thickness; HGF with keratometry, spherical equivalent, and axial length; CST5 with macular thickness; TWEAK with choroidal thickness). However, none remained significant after FDR correction for multiple comparisons (all q > 0.25). These findings are considered hypothesis-generating.ConclusionThe upregulation of HGF, CST5, CCL19, TWEAK, CXCL5, and DNER in the aqueous humor of patients with high myopia and their specific correlations with ocular parameters suggest a new molecular perspective for understanding the pathogenesis of high myopia. These preliminary findings suggest that the candidate biomarkers may have potential relevance to high myopia progression; however, independent validation in larger cohorts is needed. Introduction Myopia is an increasingly serious global eye health problem, with an incidence rate of 0.9–3.1% in Asian countries. High myopia (HM) is defined as a diopter greater than –6.0 D or axial length exceeding 26 mm, and it is a leading cause of irreversible low vision and blindness.1,2 It is projected that by 2050, the global population with myopia and high myopia will reach 4.8 billion and 0.9 billion, respectively.3 The typical fundus lesions of high myopia include leopard-spot fundus, lacquer cracks, retinal and choroidal atrophy, choroidal neovascularization, macular atrophy, posterior scleral staphyloma, and retinopathy.4 Currently, due to limited understanding of the pathogenesis of high myopia, no pharmacological interventions are available to prevent or treat myopia-related retinal degeneration.5 Therefore, elucidating the pathogenesis of high myopia has become an urgent research priority.6 Maintaining intraocular immune homeostasis is essential for normal visual function; once this balance is disrupted, it may contribute to axial elongation and myopia progression.7 Accumulating evidence suggests that the unique intraocular microenvironment plays a pivotal role in driving the development of various ocular diseases. Inflammation disrupts intraocular immune homeostasis by upregulating pro-inflammatory cytokines and matrix metalloproteinase-2 (MMP-2), thereby accelerating extracellular matrix degradation and triggering scleral remodeling—a process driven by active and dynamic pathophysiological responses including hypoxia, metabolic reprogramming, immune cell infiltration, and fibroblast phenotypic transformation—ultimately leading to axial elongation.8 Excessive axial elongation can trigger choroidal microvascular dysfunction, tissue hypoxia, and breakdown of the blood–retinal barrier, resulting in elevated levels of inflammatory cytokines in the aqueous humor and vitreous, which may establish a self-perpetuating vicious cycle[8]. In patients undergoing cataract surgery with high myopia, elevated expression levels of inflammatory cytokines (such as MCP-1, IL-8, and IL-6) and immune-related factors (such as ICAM and VCAM) have been observed in the aqueous humor, along with decreased concentrations of VEGF and TGF-β1, suggesting a mild pro-inflammatory state in highly myopic eyes.9 High myopia is considered an inflammation-related disease, and highly myopic eyes are believed to have an inflammatory intraocular microenvironment, which makes aqueous humor a practical proxy for intraocular immune status.10 The aqueous humor inflammatory profile may also influence axial elongation, highlighting its potential value as a biomarker reservoir for monitoring the progression and complications of high myopia.11 Quantitative assessment of ocular inflammation has been successfully applied in other chronic anterior segment disorders. For example, in pterygium – a fibrovascular proliferative condition – objective redness grading using software (PRGS) reliably captures inflammatory phenotypes12,13 and correlates with functional visual changes.14 These approaches underscore the value of systematic, reproducible inflammatory profiling, which we now extend to the molecular level in high myopia using the Olink platform. With the deepening of research on the immune microenvironment, inflammatory factors have gradually gained attention as key biomarkers of myopia. The Olink proteomics platform has been predominantly applied to conditions such as diabetic retinopathy, while its use in studying the intraocular microenvironment in myopia, especially high myopia, remains limited.15 In the present study, we collected aqueous humor samples from cataract patients with and without high myopia and performed systematic immunoproteomic profiling using an antibody-mediated high-throughput proteomics approach (Olink Target 96 Inflammation Panel). We hypothesized that, compared with non-myopic controls, patients with high myopia would exhibit a distinct inflammatory protein signature in the aqueous humor, and that these proteins may correlate with axial length, choroidal thickness, and other ocular parameters. This study aims to characterize the immunoproteomic features of the aqueous humor in relation to the intraocular microenvironment and to identify potential biomarkers for predicting the onset and progression of high myopia. Methods Study Subjects This study is a cross-sectional study designed to preliminarily investigate the characteristics of the intraocular inflammatory microenvironment in high myopia. Given the exploratory nature of this study and the difficulty in obtaining paired aqueous humor samples, a formal power analysis was not performed. The sample size was based on feasibility and is consistent with previous studies in aqueous humor.16 A total of 20 patients undergoing cataract surgery were enrolled and divided into two groups: the high myopia group (HM group, n=12) and the non-high myopia control group (Control group, n=8). The study protocol was approved by the Ethics Committee of Beijing Friendship Hospital, Capital Medical University (2026-P2-090-02). This study was conducted in accordance with the Declaration of Helsinki. Written informed consent regarding the use of their clinical data and biological samples (aqueous humor) was obtained from all participants prior to surgery. For patients scheduled to undergo cataract surgery between January 1, 2025 to February 1, 2026, the inclusion criteria were as follows: High myopia group: axial length ≥ 26.0 mm; non-high myopia (control) group: axial length between 21.0 and 24.0 mm17 (To avoid interference from the refractive status caused by cataract type, such as index myopia due to nuclear opacity, refractive error was not used as a grouping criterion in this study); All patients were scheduled to undergo phacoemulsification combined with intraocular lens implantation. All patients were graded for cataract using the Lens Opacities Classification System III (LOCS III). Only patients with nuclear opalescence (NO) grade 2–4 were included. To minimise the potential confounding effect of refractive index changes caused by nuclear sclerosis on axial length grouping, patients with predominantly cortical or posterior subcapsular cataracts were excluded. The exclusion criteria were as follows: Patients with posterior staphyloma, lacquer cracks, choroidal neovascularization, or macular atrophy on fundus examination or OCT were excluded from the HM group. Therefore, the HM group in this study represents early, uncomplicated high myopia. Coexistence of other fundus diseases that could affect the composition or detection results of the aqueous humor, such as age-related macular degeneration, diabetic retinopathy, retinal vein occlusion, uveitis, glaucoma, or a history of ocular trauma; Poorly controlled systemic diseases, including hypertension (systolic blood pressure >160 mmHg or diastolic blood pressure >100 mmHg), diabetes (glycosylated hemoglobin HbA1c > 7.5%), autoimmune diseases, or active infections; Use of ocular or systemic medications that could affect inflammatory or immune markers (eg, glucocorticoids, non-steroidal anti-inflammatory drugs, immunosuppressants) within three months prior to surgery; A history of prior intraocular surgery. Clinical Examination and Data Collection All patients underwent standardized ophthalmic examinations. Best-corrected visual acuity (BCVA) was measured using a standard logarithmic visual acuity chart and converted to logMAR units for recording. Anterior segment and fundus examinations were performed in detail using a slit-lamp microscope combined with a +90D fundus lens. Axial length was measured using an optical interference biometric device (IOLMaster 700®, Carl Zeiss Meditec, Germany), and the average value of five consecutive measurements was recorded. Subfoveal choroidal thickness was measured using spectral-domain optical coherence tomography (TOPCON DRI OCT Triton, Tokyo, Japan) with the enhanced depth imaging mode. Basic data, including patient age, sex, medical history, and medication history, were also collected. Acquisition of the Aqueous Humor Samples All aqueous humor samples were collected under sterile conditions at the beginning of cataract surgery, prior to the first incision. The procedure was as follows: after routine disinfection and draping, the conjunctival sac was irrigated with 5% povidone-iodine solution and allowed to dwell for at least 30 seconds, followed by thorough rinsing with sterile physiological saline. Under an operating microscope, anterior chamber paracentesis was performed at the corneal limbus using a 1 mL insulin syringe fitted with a 30-gauge needle, and approximately 100–150 µL of aqueous humor was slowly aspirated. The sample was immediately transferred to a pre-chilled, enzyme-free, sterile EP tube. Within 30 minutes of collection, the sample was rapidly frozen in liquid nitrogen and subsequently transferred to a –80°C ultra-low temperature freezer for long-term storage until batch analysis. Repeated freeze-thaw cycles were avoided throughout the process. Measurement of Aqueous Proteins The Olink Target 96 Inflammation Panel was employed in this study. This assay is based on the Proximity Extension Assay (PEA) technology, enabling simultaneous quantification of 92 target proteins. In this technique, pairs of oligonucleotide-labeled antibody probes bind to their respective target proteins; when the two probes are brought into close proximity, their oligonucleotide sequences hybridize specifically. Subsequently, a proximity-dependent DNA extension reaction is initiated by DNA polymerase, generating a unique PCR target sequence. This sequence is then detected and quantified using a microfluidic real-time PCR system. Data were subjected to quality control and normalization using internal extension controls and inter-plate controls to correct for intra- and inter-assay variations. The final results are presented as Normalized Protein eXpression (NPX) values, which are relative units on a log2 scale, where higher values indicate higher protein expression levels. No sample had an intra-assay coefficient of variation >15%; therefore, no samples were excluded, and all 20 samples were included in the final analysis. Proteomic Data Analysis Normalized Protein Expression (NPX) values were Z-score transformed for further analysis. Principal component analysis (PCA) was performed using the prcomp function to preliminarily observe the separation trend between groups. For differential expression analysis, the non-parametric Mann–Whitney U-test was used to compare NPX values of each protein between the two groups. P-values were adjusted for multiple testing using the Benjamini-Hochberg method to control the false discovery rate (FDR). Differentially expressed proteins (DEPs) were defined as those with an adjusted p-value < 0.05 and |log2(fold change)| > 0.585 (corresponding to a concentration change greater than 1.5-fold). The list of gene symbols for the identified DEPs was then submitted for further bioinformatics analysis. Gene Ontology (GO) enrichment analysis was performed using the clusterProfiler R package, encompassing the three domains of biological process (BP), cellular component (CC), and molecular function (MF). KEGG pathway enrichment analysis was conducted using the WebGestalt online tool to explore key signaling pathways involving the DEPs. For protein-protein interaction (PPI) network analysis, the DEP list was submitted to the STRING database (v11.5) to obtain interaction networks, which were then visualized and mined for core modules using Cytoscape software (v3.9.1). Statistical Analysis All statistical analyses were performed using R software (version 4.2.0). Two-sided tests were conducted, and a p-value < 0.05 was considered statistically significant. Due to the small sample size and non-normal distribution of most variables (assessed by Shapiro–Wilk test), the non-parametric Mann–Whitney U-test was used for all between-group comparisons, including baseline characteristics and protein NPX values. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. P-values were adjusted for multiple comparisons using the Benjamini-Hochberg false discovery rate (FDR) method. Differentially expressed proteins (DEPs) were defined as those with an FDR-adjusted q-value < 0.05 and |log2(fold change)| > 0.585 (corresponding to a >1.5-fold change). For correlation analyses, the Benjamini-Hochberg FDR method was applied separately for each group (HM and non-HM) to account for multiple comparisons. Given the exploratory nature of the study, an FDR threshold of q < 0.25 was considered suggestive for pilot analyses. Results The Clinical Characteristic of Patients Table 1 summarizes the baseline characteristics and ocular parameters of the control (n=8) and high myopia (HM, n=12) groups. No significant differences were observed in age, gender, or eye laterality between the two groups (all p > 0.05). As expected, the HM group had significantly longer axial length, higher refractive error, flatter corneas (K1, K2, K_Mean), and deeper anterior chamber depth compared with the control group (all p < 0.05). In addition, macular center thickness and choroidal thickness were significantly were significantly lower in the HM group (p = 0.002 and p < 0.001, respectively). However, no significant differences were found in central corneal thickness or BCVA between groups (p > 0.05).Table 1Baseline Characteristics and Ocular Parameters in Control and High Myopia GroupsCharacteristicsControl Group (n=8)High Myopia Group (n=12)Test StatisticP valueSex, n (%)Fisher0.642 Female5 (62.5)9 (75.0) Male3 (37.5)3 (25.0)Age, y68.00 (67.75, 68.25)67.00 (61.50, 70.75)Z=0.1540.907Axial length, mm22.20 (21.89, 22.64)26.75 (26.22, 27.99)Z=−3.703<0.001Refractive error0.25 (0.19, 0.25)9.50 (8.69, 11.75)Z=−3.703<0.001Spherical equivalent, D46.37 (45.18, 46.94)44.24 (43.33, 44.66)Z=3.1630.002BCVA0.25 (0.11, 0.26)0.12 (0.10, 0.17)Z=1.3890.169LogMAR0.60 (0.58, 0.94)0.92 (0.77, 1.01)Z=−1.3890.169ACD, mm3.14 (2.91, 3.27)3.62 (3.35, 3.81)Z=−2.6620.009K1, D45.84 (44.77, 46.51)43.52 (42.53, 44.30)Z=3.163<0.001K2, D46.72 (45.62, 47.48)44.73 (44.19, 45.25)Z=3.0470.003K mean, D46.37 (45.19, 46.95)44.25 (43.35, 44.66)Z=3.1630.002Macular center thickness, μm276.50 (272.75, 281.25)253.00 (248.25, 272.75)Z=2.2760.025Central corneal thickness, μm550.00 (537.25, 557.50)543.50 (533.75, 563.00)Z=−0.1160.938Choroid thickness, μm241.50 (220.75, 252.25)142.00 (118.25, 174.50)Z=3.703<0.001 Table 1Baseline Characteristics and Ocular Parameters in Control and High Myopia GroupsCharacteristicsControl Group (n=8)High Myopia Group (n=12)Test StatisticP valueSex, n (%)Fisher0.642 Female5 (62.5)9 (75.0) Male3 (37.5)3 (25.0)Age, y68.00 (67.75, 68.25)67.00 (61.50, 70.75)Z=0.1540.907Axial length, mm22.20 (21.89, 22.64)26.75 (26.22, 27.99)Z=−3.703<0.001Refractive error0.25 (0.19, 0.25)9.50 (8.69, 11.75)Z=−3.703<0.001Spherical equivalent, D46.37 (45.18, 46.94)44.24 (43.33, 44.66)Z=3.1630.002BCVA0.25 (0.11, 0.26)0.12 (0.10, 0.17)Z=1.3890.169LogMAR0.60 (0.58, 0.94)0.92 (0.77, 1.01)Z=−1.3890.169ACD, mm3.14 (2.91, 3.27)3.62 (3.35, 3.81)Z=−2.6620.009K1, D45.84 (44.77, 46.51)43.52 (42.53, 44.30)Z=3.163<0.001K2, D46.72 (45.62, 47.48)44.73 (44.19, 45.25)Z=3.0470.003K mean, D46.37 (45.19, 46.95)44.25 (43.35, 44.66)Z=3.1630.002Macular center thickness, μm276.50 (272.75, 281.25)253.00 (248.25, 272.75)Z=2.2760.025Central corneal thickness, μm550.00 (537.25, 557.50)543.50 (533.75, 563.00)Z=−0.1160.938Choroid thickness, μm241.50 (220.75, 252.25)142.00 (118.25, 174.50)Z=3.703<0.001 Baseline Characteristics and Ocular Parameters in Control and High Myopia Groups Identification of Aqueous Humor Biomarkers in High Myopia Based on the Olink Inflammation Panel Using the Olink proteomics platform, we analyzed aqueous humor samples from 12 patients with high myopia cataracts (HM group) and 8 patients with age-related cataracts (control group) to identify differentially expressed proteins. A total of 92 inflammation-related proteins were quantitatively analyzed. Principal component analysis (PCA) based on the normalized protein expression values of these 92 proteins showed a clear separation between the HM and control groups, with red dots representing HM samples (n=12) and blue dots representing control samples (n=8) (Figure 1A). Hierarchical clustering analysis of the top nine differentially expressed proteins further demonstrated distinct protein expression patterns between the two groups (Figure 1B). Statistical analysis (q < 0.05) revealed that nine proteins were significantly upregulated in the HM group compared with the control group. Among these, hepatocyte growth factor (HGF) exhibited the most prominent fold change. Volcano plot analysis highlighted HGF (log2 fold change = 0.707, FDR-adjusted q = 0.005), CST5 (log2 fold change = 0.408, q = 0.005), and CCL19 (log2 fold change = 1.153, q = 0.012) as top candidate proteins with differential expression (Figure 1C and D). These findings prompted us to further investigate the functional relevance of these upregulated proteins in the pathophysiology of high myopia. Detailed statistics for the nine DEPs are shown in Supplementary Table S1.Figure 1Identification and expression characteristics of differentially expressed proteins in aqueous humor between the high myopia group and the control group. (A) Principal component analysis (PCA) showing the overall separation trend between the high myopia group (blue) and the control group (red) samples, based on the normalized expression values of 92 inflammation-related proteins. (B) Heatmap illustrating the protein expression profiles of the high myopia group (n=12) and the control group (n=8), displaying differentially expressed inflammation-related proteins. (C) Box plot showing the upregulated differentially expressed proteins between the high myopia group and the control group. * q < 0.05, ** q < 0.01 (FDR-corrected). (D) Volcano plot of all 92 inflammation-related proteins. The vertical dashed line indicates |log2FC| = 0.585, and the horizontal dashed line indicates FDR-adjusted q = 0.05. Red dots represent significantly upregulated DEPs (q < 0.05 and |log2FC| > 0.585). No proteins met the criteria for significant downregulation.A four-plot figure comparing Groups C and H for protein expression using PCA, heatmap, box plots and volcano plot.Image A′s PCA scatter plot shows Group C and H with partial separation: Group H at negative PC1, Group C at positive PC1, overlapping near PC1=0. Image B′s heatmap with hierarchical clustering shows proteins like CST5, HGF, CCL19 and CXCL5 higher in Group H. Image C′s box plots reveal Group H′s higher medians, notably for CCL19, HGF and TWEAK. Image D′s volcano plot of 92 inflammation-related proteins has CCL19 as the most extreme with log2FoldChange near 1.1. Proteins in the volcano plot align with assays in box plots and heatmap rows, indicating higher expression in Group H, consistent with PCA separation. Identification and expression characteristics of differentially expressed proteins in aqueous humor between the high myopia group and the control group. (A) Principal component analysis (PCA) showing the overall separation trend between the high myopia group (blue) and the control group (red) samples, based on the normalized expression values of 92 inflammation-related proteins. (B) Heatmap illustrating the protein expression profiles of the high myopia group (n=12) and the control group (n=8), displaying differentially expressed inflammation-related proteins. (C) Box plot showing the upregulated differentially expressed proteins between the high myopia group and the control group. * q < 0.05, ** q < 0.01 (FDR-corrected). (D) Volcano plot of all 92 inflammation-related proteins. The vertical dashed line indicates |log2FC| = 0.585, and the horizontal dashed line indicates FDR-adjusted q = 0.05. Red dots represent significantly upregulated DEPs (q < 0.05 and |log2FC| > 0.585). No proteins met the criteria for significant downregulation. Functional Annotation of Differentially Expressed Inflammation-Related Proteins in High Myopia To explore the molecular mechanisms and biological processes involved in the pathogenesis of high myopia, we performed Gene Ontology (GO) enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and protein-protein interaction (PPI) analysis of the differentially expressed proteins (DEPs). In the GO analysis, the three domains of biological process (BP), cellular component (CC), and molecular function (MF) were analyzed against the background of 92 inflammation-related proteins in the Olink panel (Figure 2A). In terms of molecular function, the DEPs were mainly enriched in receptor ligand activity, signaling receptor activator activity, cytokine activity, and cytokine receptor binding. Regarding the cellular component, the altered proteins were primarily localized in platelet alpha granules, secretory granule lumen, and membrane rafts. In terms of biological process, the DEPs were most significantly enriched in positive regulation of cytokine production, cytokine-mediated signaling pathway, and interleukin-6 production. Subsequently, the GO enrichment scatter plot displayed the top 15 terms with the highest Rich Factor, among which “receptor ligand activity”, “signaling receptor activator activity”, “cytokine activity”, and “cytokine receptor binding” were significantly enriched (Figure 2B). KEGG pathway analysis revealed that the DEPs were mainly involved in immune system regulation and signal transduction. The most significantly enriched pathway was cytokine-cytokine receptor interaction. In addition, the chemokine signaling pathway and pathways associated with chronic inflammatory diseases, such as malaria and rheumatoid arthritis, were also significantly enriched (Figure 2C). Finally, we constructed and analyzed a protein-protein interaction network module with strong correlations to explore the interactions among the inflammation-related proteins with altered expression (Figure 2D). In the PPI network, tumor necrosis factor-like weak inducer of apoptosis (TWEAK), CXCL5, and CCL19 exhibited the highest degree scores and the most connections with other proteins, indicating that they play important roles in high myopia.Figure 2GO and KEGG enrichment analysis and PPI network analysis of differentially expressed proteins. (A) GO enrichment bar plot: The x-axis represents the GO functional categories, and the y-axis represents the number of proteins enriched in each category. (B) GO enrichment scatter plot: Bubble color indicates the significance level of enrichment for GO terms. (C) KEGG enrichment bar plot: The y-axis lists the significantly enriched KEGG pathways, and the x-axis represents the number of proteins enriched in each pathway (Rich Factor ≤ 1). (D) Protein-protein interaction network of potential biomarkers constructed using the STRING database, with edge width representing the combined score. PPI: protein-protein interaction.A composite figure with two bar graphs, a bubble scatter plot and a protein interaction network.Image A shows a bar graph with GeneRatio (1.8-6.8) on the x-axis and GO terms on the y-axis. Key terms: receptor ligand activity (6.8), signaling receptor activator activity (6.6), cytokine activity (6.2). Neuronal cell body is around 3.0, biological processes near 4.8 and 5.8. Image B has a bubble scatter plot with GeneRatio on the x-axis and -log10 adjusted p value on the y-axis. Notable points: GO:0048018 (6.6, 6.8), GO:0005125 (5.6, 6.6). A legend categorizes GO terms into BP, CC, MF, with a size legend for Count (1-6). Image C displays a bar graph with Count (0-5) on the x-axis and pathways on the y-axis. Cytokine-cytokine receptor interaction has the longest bar (5), followed by Malaria and Rheumatoid arthritis (2 each). A p.adjust legend shows values 0.04, 0.08, 0.12. Image D features a protein-protein interaction network with nodes like CXCL5, IL33, TNF, highlighting connections around CXCL5, TNFSF12, TNF, CCL19. GO and KEGG enrichment analysis and PPI network analysis of differentially expressed proteins. (A) GO enrichment bar plot: The x-axis represents the GO functional categories, and the y-axis represents the number of proteins enriched in each category. (B) GO enrichment scatter plot: Bubble color indicates the significance level of enrichment for GO terms. (C) KEGG enrichment bar plot: The y-axis lists the significantly enriched KEGG pathways, and the x-axis represents the number of proteins enriched in each pathway (Rich Factor ≤ 1). (D) Protein-protein interaction network of potential biomarkers constructed using the STRING database, with edge width representing the combined score. PPI: protein-protein interaction. Correlations Between Cytokine Levels and Ocular Parameters Spearman correlation analysis was performed to explore associations between aqueous humor cytokine levels and ocular parameters, with results presented separately for the HM and non-HM groups (Figure 3). A total of 117 Spearman correlations (9 proteins × 13 ocular parameters) were tested in each group. After FDR correction for multiple comparisons, only the association between DNER and central corneal thickness in the HM group remained suggestive (q < 0.25); all other nominally significant associations (uncorrected p < 0.05) did not survive correction and are considered hypothesis-generating. These exploratory findings provide preliminary insights into possible cytokine-ocular parameter associations. However, given the limited sample size and multiple comparison correction, independent validation is necessary. See Supplementary Table S2 for full correlation results.Figure 3Correlation analysis between aqueous humor cytokine levels and ocular parameters in the high myopia (HM) and non-high myopia (non-HM) groups. Each scatter plot shows the Spearman correlation coefficient (r), raw p-value (uncorrected), and FDR-adjusted q-value. (A) In the HM group, DNER was positively correlated with central corneal thickness (q = 0.215, reaching the suggestive threshold). (B–H) The remaining nominally significant associations (raw p < 0.05) had q-values > 0.25 (see panels B–H). All associations with q > 0.25 are hypothesis-generating and presented for exploratory illustration.A set of eight scatter plots showing cytokine levels versus ocular parameters in two groups.Image A: Scatter plot of central corneal thickness vs. DNER. Non-HM: r=0.467, P=0.05; HM: r=0.799, P=0.002. Image B: Central corneal thickness vs. TWEAK. Non-HM: r=0.036, P=0.5; HM: r=0.669, P=0.017. Image C: Macular center thickness vs. CXCL9. Non-HM: r=0.143, P=0.5; HM: r=0.627, P=0.029. Image D: Spherical equivalent vs. HGF. Non-HM: r=0.790, P=0.020; HM: r=0.294, P=0.5. Image E: Axial length vs. HGF. Non-HM: r=0.833, P=0.010; HM: r=0.119, P=0.5. Image F: K1 vs. HGF. Non-HM: r=0.810, P=0.015; HM: r=0.336, P=0.5. Image G: Macular center thickness vs. CST5. Non-HM: r=0.714, P=0.047; HM: r=0.382, P=0.5. Image H: Choroid thickness vs. TWEAK. Non-HM: r=0.762, P=0.028; HM: r=0.378, P=0.5. Correlation analysis between aqueous humor cytokine levels and ocular parameters in the high myopia (HM) and non-high myopia (non-HM) groups. Each scatter plot shows the Spearman correlation coefficient (r), raw p-value (uncorrected), and FDR-adjusted q-value. (A) In the HM group, DNER was positively correlated with central corneal thickness (q = 0.215, reaching the suggestive threshold). (B–H) The remaining nominally significant associations (raw p < 0.05) had q-values > 0.25 (see panels B–H). All associations with q > 0.25 are hypothesis-generating and presented for exploratory illustration. Discussion Although the pathological mechanism of high myopia is extremely complex, accumulating evidence suggests that inflammation and immune dysregulation play a pivotal role in its onset and progression. Therefore, the identification of sensitive and reliable inflammatory protein biomarkers in the aqueous humor has become a key step in optimizing disease management. In the present study, we utilized the Olink proteomics platform to profile the expression of 92 inflammation-related proteins in the aqueous humor of patients with high myopia (HM) and non-high myopia controls. The panel encompassed pro- and anti-inflammatory cytokines, chemokines, growth factors, and factors involved in acute inflammation, immune response, angiogenesis, fibrosis, and endothelial activation. The results revealed a total of nine significantly upregulated differentially expressed proteins (DEPs). GO and KEGG enrichment analyses demonstrated that the pathogenesis of high myopia is a multifactorial and complex biological process, with the cytokine-cytokine receptor interaction pathway identified as the most core enriched pathway, suggesting that this pathway may occupy an important position in the pathogenesis of high myopia. At the molecular function level, DEPs were mainly enriched in terms such as receptor ligand activity and cytokine activity, indicating that these proteins may primarily participate in immune regulation as signaling molecules. At the cellular component level, enrichment of platelet alpha granules suggested that they may release inflammatory factors through degranulation. At the biological process level, positive regulation of cytokine production and interleukin-6 production were the most significant enrichment results, indicating that these proteins may function by driving the inflammatory cascade. PPI analysis of all DEPs revealed that TWEAK, CXCL5, and CCL19 were closely interconnected in the PPI network and were highly expressed in the high myopia group, suggesting potential synergistic effects that warrant further investigation in high myopia. A total of nine significantly upregulated proteins were identified in the HM group, among which HGF, CST5, and CCL19 exhibited the most prominent differential expression. Based on this, we further analyzed the relationships between these inflammation-related proteins and ocular clinical parameters. Among the upregulated proteins, hepatocyte growth factor (HGF) showed the most significant fold change. HGF is a factor known to regulate cell proliferation, migration, and tissue remodeling; in microenvironments with chronic, persistent injury, aberrant HGF signaling may drive mesenchymal cell activation and abnormal extracellular matrix (ECM) deposition, thereby exacerbating the fibrotic process.18 In the field of ophthalmology, research on HGF has mainly focused on corneal wound healing, vascular leakage, and neovascularization in diabetic retinopathy.19,20 The present study found that HGF correlated with axial length, corneal curvature (K1), and spherical equivalent in the non-myopia group, whereas this correlation disappeared in the high myopia group. This phenomenon may reflect a functional transition of HGF at different stages of myopia. HGF may be a key factor promoting the protection of retinal structure and function, primarily involved in physiological ocular growth and corneal curvature regulation. Previous studies have also supported the role of HGF in maintaining homeostasis in normal scleral and corneal tissues.21,22 However, in highly myopic eyes, where axial length has entered a phase of pathological over-elongation, HGF expression may be overshadowed by other stronger pathological factors (such as TWEAK, CXCL5, etc). At this point, the correlation between HGF and axial length is disrupted, replaced by correlations of these inflammatory factors with posterior segment parameters such as choroid and macula. This suggests that under the pathological state of high myopia, the regulatory network of the intraocular microenvironment might undergo remodeling, potentially with the dominant role of HGF being partially replaced by other factors. Consistent with previous studies, Lu et al23 observed significantly elevated HGF levels in the aqueous humor of high myopia patients, which were positively correlated with axial length and the severity of posterior capsular opacification (PCO), suggesting a potential role for HGF in scleral remodeling that deserves further study in myopic eyes. In addition to HGF, this study also identified significantly upregulated proteins such as CST5, CCL19, TWEAK, DNER, and CXCL5. Their coordinated upregulation suggests the presence of extensive inflammatory and immune response activation within highly myopic eyes, potentially driving scleral remodeling and retinal choroidal degeneration. Consistent with previous studies, our HM group exhibited significantly thinner choroidal and macular thickness compared with controls, reflecting progressive choroidal vascular atrophy, reduced blood flow, and tissue hypoxia.24 Among the identified upregulated proteins, CST5, also known as cystatin D, is a cysteine protease inhibitor that regulates intracellular protein degradation and metabolic processes by inhibiting specific protease activities.25 CCL19 is a chemokine that plays an important role in inflammatory responses, mediating the migration and homing of dendritic cells and T lymphocytes to lymphoid tissues and peripheral sites.26 Current research is consistent with the hypothesis that exosomes derived from LECs can lead to decreased CCL19 expression levels in microglia, reducing ocular neovascularization by limiting the activation and aggregation of microglia.27 These two factors have been sparsely studied in the ophthalmic field, and no research has yet explored the association between CST5 or CCL19 and myopia. The present study reports for the first time the upregulation of CST5 and CCL19 in the aqueous humor of high myopia patients. The upregulation of these chemokines may indicate activation of immune-related pathways during myopia progression, consequently leading to chronic low-grade inflammation in highly myopic eyes. TWEAK (or Tumor Necrosis Factor Superfamily Member 12, TNFSF12) belongs to the tumor necrosis factor superfamily and activates multiple signaling pathways by binding to the fibroblast growth factor-inducible 14 (Fn14) receptor, participating in proliferation, migration, inflammation, and angiogenesis.28,29 High levels of TWEAK can induce cardiac remodeling and promote inflammatory responses and fibroblast proliferation.30 In this study, TWEAK showed nominal significance with central corneal thickness in the high myopia group and with choroidal thickness in the non-high myopia group, raising the possibility of an involvement in corneal stromal remodeling and choroidal vascular regulation. Similar to our findings, a Mendelian randomization analysis indicated a causal association between TWEAK and myopic refractive error, with each 1 pg/mL increase in TWEAK associated with a reduction of 0.104 D (−0.152 to −0.055; P=2.878×10−5) in myopic refractive error.31 Although the mechanism of action of TWEAK in the eye remains unclear, our results may suggest that it may participate in the scleral remodeling process, thereby influencing the development of myopic refractive error. DNER is a transmembrane protein that mediates Notch signaling through intercellular interactions in neurons and is involved in the maturation of Bergmann glial cells.32 Recent studies have found that DNER is also expressed in macrophages and is elevated in the lung tissue of patients with chronic obstructive pulmonary disease and mouse models; in pro-inflammatory recruited macrophages, DNER regulates IFNγ secretion through the non-canonical Notch pathway.33 In the present study, we observed for the first time a positive correlation between DNER and central corneal thickness in the high myopia group (suggestive association under the exploratory threshold, q = 0.215). However, this finding requires independent validation. If confirmed, it raises the possibility that DNER might be involved in corneal remodeling in high myopia, potentially through regulating local corneal immune cells or the Notch signaling pathway in corneal stromal cells — a hypothesis that requires direct experimental testing. CXCL5 is a homolog of IL-8 and is released by inflammatory cells upon activation by IL-1 or TNFα. By interacting with CXC chemokine receptor 2 (CXCR2), it exerts a potent chemotactic effect on neutrophils, playing a critical role in leukocyte infiltration in inflammatory diseases, and is also involved in angiogenesis and connective tissue remodeling.34,35 In the present study, CXCL5 was nominally correlated with macular center thickness in the high myopia group (uncorrected p = 0.029). This raises the possibility of an early molecular event in myopic maculopathy, which remains speculative. CXCL5, acting as a chemokine, recruits immune cells to the macular area, triggering chronic inflammation and ultimately leading to macular thinning and atrophy. Notably, KEGG enrichment analysis revealed that the chemokine signaling pathway was one of the significantly enriched pathways among the DEPs, providing pathway-level evidence for the involvement of CXCL5 in intraocular immune cell recruitment and tissue remodeling. This aligns with the pro-fibrotic role of CXCL5 in proliferative vitreoretinopathy. Studies have shown that CXCL5 expression is significantly elevated in patients with traumatic proliferative vitreoretinopathy (TPVR) and correlates with final best-corrected visual acuity (BCVA-L), suggesting that inhibition of CXCL5 may be a potential therapeutic target for improving TPVR.36 Beyond molecular profiling, objective clinical grading of ocular inflammation – such as the Pterygium Redness Grading Software (PRGS) validated by Hilmi et al12 and Che Azemin et al13– has proven reliable for anterior segment diseases. In high myopia, similar quantitative tools that combine proteomic biomarkers with clinical imaging (eg, OCT-based inflammatory indices) could be developed. Moreover, the functional impact of different pterygium types on vision14 encourages us to investigate whether specific proteomic signatures in HM correspond to distinct patterns of visual function decline. There are several limitations to acknowledge. The small sample size (n=20) limits statistical power and generalizability, so the results are exploratory and hypothesis‑generating. Without an independent validation cohort, we cannot make strong claims about biomarker performance. Also, because all patients had cataracts, the findings might not be generalizable to high myopia patients without cataract. The cross‑sectional design and the use of patient‑derived aqueous humor samples mean that all findings are correlational, and causality cannot be inferred. While we used FDR correction for differential expression, the correlation analyses are still exploratory, and some associations could be due to chance. In addition, the modest fold changes may limit biological significance and require functional validation using orthogonal methods. Future experiments using cell lines (eg, scleral fibroblasts) or animal models (eg, form‑deprivation myopia models) are needed to further validate whether the inflammatory proteins identified in this study (eg, HGF, TWEAK, CXCL5) are directly involved in scleral remodeling, choroidal thinning, and axial elongation. Conclusion In conclusion, this exploratory study, based on the Olink proteomics platform, provides the first application of a 92-plex inflammation panel to the aqueous humor of patients with high myopia. Nine significantly upregulated differentially expressed proteins were identified. GO and KEGG enrichment analyses suggest that these proteins are enriched in cytokine activity, chemokine signaling pathways, and immune-inflammatory responses. Exploratory correlation analyses suggested potential associations with several ocular parameters (eg, axial length, corneal curvature, choroidal thickness, and macular thickness), though most were not robust after multiple comparison correction. These preliminary findings offer initial molecular insights into the pathogenesis of high myopia. However, all findings must be validated in an independent, larger cohort using orthogonal methods (eg, ELISA, SIMOA), and longitudinal studies are needed to assess causality.