Response profiles to a controlled adverse desiccating environment based on clinical and tear molecule changes

Purpose: To investigate response profiles in the lacrimal functional unit of dry eye disease (DED) and healthy volunteers after exposure to a controlled adverse desiccating environment (CADE) by identifying groups of individuals with similar clinical and molecular changes. Methods: Clinical parameters and tear molecule levels of 20 mild-moderate DED patients and 20 healthy volunteers were evaluated pre- (baseline) and post-CADE exposure. Clustering based on relative change from baseline values was used to identify response profiles. One-vs-all logistic regression was used to identify baseline predictors for response clusters. Results: Four response profiles were identified. Cluster 1: tear break-up time (TBUT) decrease and matrix metalloproteinase 9 (MMP-9) increase. Cluster 2: marked increase in corneal staining, up-regulation of both MMP-9 and interleukin (IL)-6 levels, and down-regulation of epithelial growth factor (EGF). Cluster 3: increase in fractalkine, vascular endothelial growth factor (VEGF), MMP-9, IL-6, IL-8, IL-1 receptor antagonist (IL-1Ra) and RANTES (regulated on activation, normal T expressed and secreted) tear levels; and increased corneal staining and decreased TBUT and phenol red thread scores. Cluster 4: decreased single-item score dry eye questionnaire (SIDEQ) scores and increased corneal staining. Predictive models using baseline variables found that cluster membership depended on: corneal and conjunctival staining, SIDEQ score, interferon gamma-induced protein (IP)-10, VEGF, and IL-1Ra concentrations. Conclusions: The response of both mild-moderate DED and healthy asymptomatic individuals to environmental stress (CADE) can be predicted based on baseline (pre-exposure) clinical and tear molecular parameters. Thus, identifying individuals with a predictable response could improve patient enrollment in DED clinical trials.