Name

225 - Epidémiologie des kératites neurotrophiques : prévalence, étiologies, retentissement et prise en charge thérapeutique

To evaluate the prevalence, etiologies, outcomes and clinical management of neurotrophic keratitis in a tertiary referral eye hospital.

This is a retrospective, observational, monocentric, cohort study led in a tertiary referral eye hospital (“Fondation Ophtalmologique Adolphe de Rothschild”, Paris, France). All medical records with diagnosis coding for neurotrophic keratitis from November 2009 to October 2017 were reviewed. Diagnosis of neurotrophic keratitis was accepted if corneal hypoaesthesia or anesthesia associated with epithelial alteration was noticed in the medical record. Etiologies, initial visual acuities (logMAR), patient’s age, ulcer grade using Mackie classification at initial presentation, total number of consultations, secondary infection occurrences, healing rates and times, needs of hospitalization and time, delay in diagnosis, estimation of total cost and treatment modalities used were recorded.

Within the target population of 305,351 patients, 354 eyes of 335 patients developed neurotrophic keratitis. The incidence of neurotrophic keratitis was 5.79/10,000. According to the Mackie classification, 116 eyes were categorized as stage 1 (32.8%), 108 eyes as stage 2 (30.5%), 115 eyes as stage 3 (32.5%), 14 eyes (3.94%) as pre-perforated or perforated state. Neurotrophic keratitis was of multifactorial origin in 121 eyes (34.2%). Etiologies included herpetic keratitis sequelae (114 eyes, 32.11%), iatrogenic causes (113 eyes, 31.83%), central causes (98 eyes, 27.61%), chronic ocular surface diseases (62 eyes, 17.46%), diabetes (37 eyes, 10.42%) and other rare conditions (9 eyes, 2.54%). 118 eyes (33.3%) needed surgical procedures including overlay (82 cases, 23.2%) or inlay (43 cases, 12%) amniotic membrane transplantation. Success rates of amniotic membrane transplantation, autologous serum and matrix regeneration therapy were respectively 59.8%, 44.7%, 23.3%, with a mean healing time of 18.4, 17.3 and 20.3 days. Among 354 eyes, 220 were effectively confirmed as neurotrophic keratitis without delay in diagnosis (62.1%). Mean time delay in diagnosis among the remaining eyes was 38.8 days, CI 95% = [48.0; 29.6]. In 160 cases (45.2%), topical antibiotics were given. 23 eyes (6.48%) were complicated by proven secondary infection. 104 eyes (29.4%) showed an improvement in visual acuity with treatment, 75 eyes (21.2%) showed no change in final visual acuity, and 50 eyes (14.1%) ended up with worse visual acuity than at presentation. Predictive factors for low final visual acuity were low values of initial affected eye visual acuity, age and lagophtalmos (respectively Spearman-r = 0,782, 0,344, 0,332, p < 0.05). Concerning the affected eye visual acuities change, we found significant correlation with the initial affected eye visual acuity (logMAR) and initial Mackie stage (respectively Spearman-r = -0,495 and 0,407, p < 0,0001).

Our study supported herpetic keratitis sequelae to be the most common cause, followed by iatrogenic etiologies, central neurological etiologies and diabetes. Retrospective analysis of treatment success showed that the most effective treatment for corneal neurotrophic ulcers (Mackie stage 2 or 3) seemed to be amniotic membrane transplantation. In opposite, matrix regeneration therapy seemed to be poorly effective.

To our knowledge, we reported the first large epidemiological cohort study concerning neurotrophic keratitis. This rare condition is poorly-known by physicians and is associated with poor visual outcomes and a challenging clinical management.