Prevalence of neurotrophic keratopathy in patients with chronic ocular graft-versus-host disease

Purpose: To determine the prevalence, clinical characteristics, and risk factors associated with neurotrophic keratopathy (NK) in patients with chronic ocular graft-versus-host disease (oGVHD).

Design: Retrospective cohort study.

Methods: We performed a chart review of patients diagnosed with chronic oGVHD between January 2015 and December 2018 at a single academic institution and recorded demographic data, systemic and ocular comorbidities, history of hematologic malignancy, transplant characteristics, oGVHD severity scores, and adnexal and ocular examination findings. We determined the prevalence of NK and clinical characteristics associated with NK in these patients. A multivariate logistic regression analysis was performed to determine the risk factors associated with NK in these patients.

Main outcome measure: Prevalence of NK in chronic oGVHD.

Results: We identified 213 patients diagnosed with chronic oGVHD following hematopoietic stem cell or bone marrow transplantation from our electronic patient database, and the prevalence of NK was 14%. The mean age of oGVHD patients with NK was 62.6 ± 12.9 years; 48% were women, 19 had unilateral NK, and ten had bilateral NK. In the cohort, 56%, 20%, and 24% eyes of the patients had grades 1, 2, and 3 of NK, respectively. The mean time to diagnose NK after transplantation was 52.9 ± 45.4 months. oGVHD patients diagnosed with NK had a significantly higher NIH oGVHD severity score (p = 0.04) and a lower corneal sensation score (p = 0.0001) than those without NK. Our analyses showed a significantly higher CFS score (p = 0.01) and a trend toward lower Schirmer test scores (p = 0.16) and tear break-up times (p = 0.08) in oGVHD patients with NK. Additionally, we observed a significantly higher prevalence of persistent epithelial defect (p = 0.0001), corneal ulceration (p = 0.0001), and corneal perforation (p = 0.005) in oGVHD patients diagnosed with NK. A logistic regression analysis to determine factors associated with NK showed that a higher NIH oGVHD score (odds ratio [OR] = 2.03, p = 0.026) and history of cataract surgery (odds ratio [OR] = 5.03, p = 0.001) are significant risk factors for NK in oGVHD patients.

Conclusions: The prevalence of NK in chronic oGVHD patients was 14% during the study period. Our analysis shows that oGVHD patients with a higher NIH oGVHD severity score and previous history of cataract surgery are at a higher risk of developing NK and may develop severe sequelae such as persistent epithelial defect or corneal ulceration.