Association of Tear Cytokine Ratios with Symptoms and Signs of Dry Eye Disease: Biomarker Data from the Dry Eye Assessment and Management Study

Purpose: To assess the relationship between tear inflammatory cytokine ratios (CRs) and signs and symptoms of dry eye disease (DED) to investigate the possible use of tear CRs, which may better address the complexity of cytokine interactions than absolute cytokine levels, as a DED biomarker. Methods: Tear concentrations of IL-1b, IL-6, IL-8, IL-10, IL-17A, IFN-g, and TNF-a were measured using standardized procedures, as were DED signs (corneal and conjunctival staining scores, tear break-up time, Schirmer test, Meibomian gland plugging, tear osmolarity, composite sign severity score) and symptoms [Ocular Surface Disease Index (OSDI)]. Ratios between pro-inflammatory (IL-1b, IL-8, IL-17A, IFN-g, and TNF-a) and anti-inflammatory (IL-10) cytokines were calculated. Given its opposing roles in inflammation, IL-6 was tested as both a pro- and anti-inflammatory cytokine. Correlations between CR and DED symptoms and signs were calculated using Spearman correlation coefficients (rho). Results: At baseline, 131 patients, 80.2% female and mean age 54.2 years (SD 14.1, range 20-82), from 10 sites of the Dry Eye Assessment and Management study had sufficient tear volumes >= 4 mu L for analysis. IL-6:IL-10, IL-8:IL-10, and TNF-a:IL-10 had some significant correlations, mostly with conjunctival or corneal staining or the composite sign severity score (IL-8:IL-10 and conjunctival staining: rho = 0.19, p = 0.03; IL-6:IL-10 and corneal staining: rho = 0.31, p < 0.001; IL-8:IL-10 and corneal staining: rho = 0.21, p = 0.01; IL-6:IL-10 and composite sign severity score: rho = 0.26, p = 0.003; IL-8:IL-10 and composite sign severity score: rho = 0.26, p = 0.003; TNF-a:IL-10 and corneal staining: rho = 0.29, p < 0.001; TNF-a:IL-10 and Schirmer test: rho = -0.23, p = 0.009). CRs had no significant correlation with DED symptoms. All significant correlations using IL-6 in the denominator (anti-inflammatory) produced counterintuitive results based on clinical understanding of the disease. Conclusions: Pro- to anti-inflammatory CR was weakly correlated with some DED signs and not with symptoms, as measured by OSDI. Future studies in different dry eye populations are needed and should address sampling biases and tear collection techniques.