The aim of this study was to investigate the relationship between ocular surface inflammation and corneal sub-basal nerve in meibomain gland dysfunction (MGD) patients.

From August 2017 to November 2018, 38 MGD patients (76 eyes) and 25 healthy controls (50 eyes) in Beijing Tongren Eye Center, Beijing Tongren Hospital affiliated to Capital Medical University were collected. Among of them, there were 22 male (44 eyes) and 41 female (82 eyes) with an average age of (41.7±10.4) years-old (ranged from 18 to 54 years-old). All subjects were checked corneal fluorescein staining, tear film break-up time (TBUT), lipid layer thickness measurement (LLT), Schirmer Ⅰ test, and conducted ocular surface disease index (OSDI) by questionare, lid margin score using slit-lamp microscope, and DC density, rate of meibomian gland loss using infrared photography for meibomian gland, the dendritic cell (DC) density, nerve number, nerve density, average nerve length, maximum nerve length, minimum nerve length using in vivo confocal microscopy (IVCM). The ocular surface indicators were expressed as ±s, and the comparison of those between MGD group and control group was performed using the independent sample t test. Corneal fluorescein staining, lid margin score were expressed as median and quartile, and compared by rank-sum test. Pearson correlation was used for correlation analysis.

The OSDI score, TBUT, score of corneal fluorescein staining, Schirmer Ⅰ value, rate of meibomian gland loss, lid margin score of MGD patients were (41.16±21.17), (3.13±2.35) s, 0.5 (0, 1) scores, (7.00±5.49) mm, (24.73±11.65)%, (56.77±23.64) nm, 3.0(2.25, 3.50) scores, respectively. There were non-significant difference in LLT between them (t=-1.171, P>0.05), significant difference in other indices between them expect LLT (t=10.321, -17.362, -5.801, 2.417; P<0.05); significant difference in corneal fluorescein staining and lid margin score using rank-sum test (T=11.810, 11.532; P<0.05). After IVCM examination, DC density, nerve number, nerve density, average nerve length, maximum nerve length, minimum nerve length of MGD patients were (30.52±16.81)/mm², (15.50±6.75)/p, (16 767±6418)μm/mm², (184.77±34.61)μm, (412.83±60.47)μm, (43.30±24.67)μm, respectively with significant difference from controls (t=2.129, -2.143, -3.529, 2.188; P<0.05) expect nerve number and average nerve length (t=-1.541, 1.481; P>0.05). Using ocular surface indicators of MGD patients, nerve density was negatively associated with OSDI, score of corneal fluorescein staining, the minimum nerve length. The nerve density was positive correlated with nerve number and the maximum nerve length with significant difference (r=-0.365, -0.506, -0.536, 0.888, 0.698; P<0.05). DC density was positively correlated with score of lid margin and minimum nerve length. The nerve density, nerve number and maximum nerve length all were showed negative correlations with DC density. After Pearson correlation analysis, there was significant difference (r=0.446, 0.545, -0.440, -0.459, -0.418; P<0.05).

OSDI, score of corneal fluorescein staining, rate of meibomian gland loss, lid margin score was increased in MGD patients, while TBUT and Schirmer I value decreased. The result of IVCM suggested that there was a correlation between DC density and sub-basal corneal nerve, indicating that ocular surface inflammation may induce the sub-basal corneal nerve injury.