Purpose: To systematically review the evidence on the diagnostic accuracy and prognostic value of retinal optical coherence tomography (OCT) to detect visual acuity (VA) or visual field (VF) loss in children with a brain tumour.
Methods: PubMed, Embase and Cochrane Library databases were searched from inception to February 2021. We included studies evaluating retinal OCT and standard visual function parameters (VA and or VF) in children with a brain tumour. Two authors independently extracted data from each included study. They also assessed the methodological quality of the studies using the QUADAS-2 or QUIPS tool. The diagnostic accuracy of OCT was evaluated with receiver operating characteristic analysis, sensitivity, specificity, positive predictive value and negative predictive value. The prognostic value of OCT was evaluated with predictive measures (odds ratio).
Results: We included five diagnostic studies, with a total of 186 patients, all diagnosed with optic pathway glioma. No prognostic studies were eligible for inclusion. Included studies evaluated either retinal nerve fiber layer (RNFL) thickness or ganglion cell layer-inner plexiform layer (GCL-IPL) thickness. There was considerable heterogeneity between OCT devices, OCT protocols, visual function parameters and threshold values. Sensitivity and specificity for RNFL thickness measurement ranged from 60.0% to 100.0% and 76.6% to 100%, respectively. For GCL-IPL thickness measurement, area under the curve ranged from 0.91 to 0.98 for different diameters.
Conclusion: The literature regarding the diagnostic accuracy and prognostic value of OCT parameters in children with a brain tumour is scarce. Due to heterogeneity and a considerable risk of bias of included studies, we cannot draw solid conclusions regarding the accuracy of retinal OCT. Future research should investigate the potential of OCT as diagnostic and prognostic tool for the evaluation of the visual function and detection of visual impairment in children with any type of brain tumour.