Ophthalmological examination in neurofibromatosis type 1: a long-term retrospective analysis.

Abstract

Neurofibromatosis type 1 (NF1) is a common autosomal dominant inherited disorder. Patients are prone to develop both benign and malignant tumours of the central and peripheral nervous system. Because 15% of NF1 children develop an optic pathway glioma (OPG), patients undergo regular ophthalmological screening. The overall survival is excellent, but when visual function is decreasing or when tumour growth is observed on MRI, treatment with chemotherapy should be considered to maintain or improve visual function (Listernick et al. 2007). We performed a retrospective analysis including 1364 ophthalmological visits of 168 NF1 children (79 girls; 89 boys) between 2001 and 2013 in a single centre (University Hospital Leuven, Belgium). Forty NF1 children (20 boys; 20 girls) in this cohort had a diagnosis of OPG (23.8%). The average age of diagnosis of the OPG was 5.5 years. Sixteen patients (40%) were treated. Fifteen received chemotherapy during 16 months average according to the International Society of Paediatric Oncology–Low Grade Glioma guidelines. One older patient who underwent radiotherapy was excluded from the analysis. Both the high incidence of OPG patients in this cohort and high treatment number can be accounted for by the referral bias to a tertiary centre with a paediatric oncology department. We found a significantly worse vision and significantly higher risk (eight times) of an abnormal optic disc in the presence of an OPG compared to the patients without an OPG in the corresponding eye(s; p < 0.0001). We also found that NF1 patients without an OPG have a mean VA of 0.08 LogMar, which is comparable with that of the normal population, as is the age dependent evolution (Fulton et al. 2013) (Fig. 1). At MRI diagnosis, 46% of patients had ophthalmological symptoms. During follow-up, up to 61% were symptomatic. Five patients in our series at some time developed a visual field defect without accompanying loss of vision. Three children (7.3%) presented with precocious puberty. In 11 children, the diagnosis was made after the age of 6. Despite their older age, still, six of them were symptomatic. Table 1 presents the evolution of visual function (VA and overall visual function) as well as optic disc and tumour volume on MRI in the treated and observed groups, this confirms the previously stated poor correlation between both (Blanchard et al. 2016). Only five patients showed functional improvement (VA and/or visual field) with therapy. All had only mild to moderate, newly acquired loss of visual function at presentation. The response to treatment was absent in patients presenting with poor vision. Table 2 presents the predictive value of different clinical parameters of the ophthalmological examination to detect an OPG. We found that funduscopy and VA testing are the most important screening parameters. Visual field examination, pupillary reflexes and cover test have additional value, although their sensitivity is very low. Considering the binocular vision of all OPG patients at their final visits, according to the WHO criteria, 17% of patients were visually impaired. Only one patient was legally blind. Four groups associated with a more guarded visual prognosis have been proposed by other authors and were confirmed in our study: infants under 2 years of age, chiasmic location, intraconal lesion and gender (Opocher et al. 2006). Our study demonstrates we lack objective information about visual function in infants to allow early detection of clinical significant OPGs in this age group. Promising biomarkers tools for vision such as OCT are being evaluated (Banc et al. 2018). The overall poor treatment response found in our study confirms us to be very strict in patient selection for treatment and does us reflect on the rationale for treatment and weighing the limited benefit against the potential side-effects.