PURPOSE:

To assess corneal sensitivity changes in patients with ocular graft-versus-host disease using a non-contact and Cochet-Bonnet esthesiometer. In addition, we evaluate the association between corneal sensitivity and subbasal nerve changes and epitheliopathy in these patients.

METHODS:

In this retrospective study, the clinical data and images were evaluated for 36 patients (19 female, 17 male) who fulfilled the inclusion criteria. The analyzed data included demographic and ocular surface parameters, including best-corrected visual acuity, corneal sensitivity with non-contact (mbar) and Cochet-Bonnet (cm) esthesiometer, corneal fluorescein staining (CFS) and symptoms scores, tear volume (Schirmer-I test, mm/5'), and subbasal nerve density (µm/mm2; assessed with in vivo confocal microscopy).

RESULTS:

The mean age of the study cohort was 59.9 ± 10.5 years. The mean corneal sensitivity assessed by Cochet-Bonnet and non-contact esthesiometer was 5.9 ± 0.3 cm and 7.3 ± 2.0 mbar, respectively. The ocular surface parameters included a corneal fluorescein staining (CFS) score, as per the National Eye Institute grading scheme, of 6.9 ± 3.5, and a Schirmer-I test result of 7.5 ± 6.2 mm/5 minutes.. Total corneal subbasal nerve density was inversely associated with CFS scores (r = -0.74; P < 0.001). Moreover, similar correlations between CFS scores and main trunk (r = -0.62; P < 0.001) and branch (r = -0.59; P < 0.001) nerve densities were observed. A significant correlation was found between reduced corneal sensitivity and higher CFS scores (r = 0.66; P < 0.001). Higher pressures were correlated with lower total (r = -0.83; P < 0.001), main trunk (r = -0.62; P < 0.001), and branch (r = -0.72; P < 0.001) nerve densities. The univariate analysis showed that corneal sensitivity loss (assessed with non-contact esthesiometer) was correlated with advanced age of the patients (P = 0.049) and inversely associated with total (P < 0.001), main trunk (P < 0.001), and branch (P < 0.001) nerve densities. In addition, sensitivity loss was inversely associated with punctal occlusion (cauterization (P = 0.001) or plug placement (P < 0.001). The multivariate analysis adjusted for age and punctal occlusion confirmed the associations in the univariate analysis.

CONCLUSIONS:

In this study, we observed that corneal sensitivity loss was associated with reduced main trunk, branch, and total nerve density in patients with ocular graft-versus-host disease. In addition, a significant correlation was observed between reduced corneal nerve density, corneal sensitivity, and severity of epitheliopathy.