Is Meibomian Gland Dysfunction One Disease? Heterogeneity Among Phenotypes

Abstract

Purpose: The purpose of this study was to examine ocular surface symptoms, tear metrics, and tear cytokines by Meibomian gland dysfunction (MGD) features. Methods: Symptom questionnaires and an ocular surface evaluation were performed on 40 individuals with varied MGD signs [Meibomian gland (MG) plugging, eyelid vascularity, meibum quality, and MG dropout]. Tear proteins were extracted off Schirmer strips and analyzed for 23 human inflammation-related proteins. Statistical analysis was performed to examine associations between dry eye metrics inflammatory proteins and MGD features. Results: The study involved 40 South Florida veterans with a mean age of 61 ± 13 years; most individuals were male (95%), White (31%), and non-Hispanic (85%). MGD features differentially related to dry eye signs. Eyelid vascularity, meibum quality, and MG dropout, but not MG plugging, correlated with higher corneal staining and lower tear production. MGD features also differentially related to tear cytokines. Eyelid vascularity most closely related to inflammation with significant correlations for interferon-gamma-γ (r = 0.36, P = 0.02), interleukin-4 (IL-4) (r = 0.43, P = 0.006), IL-17A (r = 0.42, P = 0.007), matrix metalloproteinase-2 (r = 0.39, P = 0.01), C-X-C motif chemokine ligand 5 (Regulated upon Activation, Normal T-Cell Expressed and presumably Secreted [RANTES]) (r = 0.32, P = 0.04), and tumor necrosis factor α (r = 0.36, P = 0.02). The other 3 MGD signs were less related to inflammation. Multivariable models revealed IL-4 to be most closely related to eyelid vascularity (standardized β = 0.39, P < 0.0001). Conclusions: Eyelid vascularity was the MGD sign most closely related to inflammatory cytokines, suggesting that different MGD features may be driven by different pathophysiological mechanisms.