Conjunctival Matrix Metalloproteinase-9 Clinical Assessment in Early Ocular Graft versus Host Disease

PURPOSE: Graft versus Host Disease (GVHD) typically affects the ocular surface, with a presentation resembling Dry Eye Disease (DED). Although the etiopathology is not completely known, the conjunctiva might be a key site of T-cell activation. The differential diagnosis might be tricky at early stages, because of the lack of dedicated clinical and laboratory tests. To meet these needs, we evaluated the suitability of ocular surface matrix metalloproteinase-9 (MMP-9) clinical test. METHODS: Consecutive GVHD patients, referred to IRCCS San Raffaele Scientific Institute, were recruited. DED patients served as controls. MMP-9 was tested through InflammaDry immunoassay kit in both groups; Ocular Surface Disease Index (OSDI) questionnaire, tear osmolarity, fluorescein Tear Break-up Time (TBUT), corneal and conjunctival staining, and Schirmer test I were also collected. Parametric and nonparametric statistical tests were used to analyze the intergroup differences; Receiver Operating Characteristics (ROC) curve analysis was carried out to perform sensitivity and specificity evaluations. RESULTS: Forty-five GVHD and 40 DED patients were included. MMP-9 expression was significantly higher in GVHD group than in DED (84.4% vs 33%, p ≤ 0.001). Corneal and conjunctival staining scores resulted worse in GVHD than in DED (0.95 ± 1.16 vs 0.40 ± 0.63, p=0.02; 0.77 ± 0.42 vs 0.40 ± 0.49, p=0.0005, respectively). No significant differences regarded the other collected parameters. GVHD group was characterized by positive correlations between MMP-9 and conjunctival staining (rho = 0.55, p=0.0002) and between MMP-9 and OSDI (rho = 0.3, p=0.01); a faint inverse correlation was found between MMP-9 and Schirmer test (rho = −0.25, p=0.04). CONCLUSION: MMP-9 has a role in physiologic cellular remodeling; when a proinflammatory stimulus occurs, MMP-9 molecules are overreleased in the extracellular matrix. The positive expression of MMP-9 in GVHD may be interpreted as the consequence of a T-cell aggression against self-antigens and may be considered a reliable biomarker to detect ocular surface inflammation in GVHD, even in early stages of the disease.