Abstract
Purpose :
We report previously unrecognized abnormalities of retinal ganglion cell (RGC) anatomy and physiology that help explain a striking disparity in young children with neurofibromatosis type 1 (NF1) who develop optic pathway gliomas (OPGs): the degree of visual loss is inadequately predicted by tumor characteristics or treatment response. In an established animal model with developmental glial defects, we find discrete patterns of RGC loss and an immediately retrobulbar demyelinated zone, and disrupted physiology prior to tumor formation or RGC loss, and among surviving RGCs.
Methods :
In mutant and littermate transgenic mice with neurofibromin (Nf1) knocked out in glial precursor cells (NF1-OPG), we compared visual behavior (optomotor responses, OMR), in vivo anatomy (optical coherence tomography, OCT), in vitro multielectrode recording of spontaneous and light-evoked RGC activity, and histologic analysis of RGCs, optic nerves and chiasms.
Results :
NF1-OPG mice proliferate atypical glial cells and develop OPGs by P15-21, with progressive loss of RGCs. Between 1 and 7 months of age, NF1-OPG mutants had moderate to severe loss of OMR sensitivity, thinning of the central retina on OCT, and diffuse or regional loss of RGCs (either discrete sectors or focal patches). Even in neighboring regions with normal RGC density, many cells either had no recordable activity or had higher spontaneous discharge rates and weaker or suppressive light responses. In many cases the retrobulbar optic nerve was severely demyelinated – even when the remaining nerve and chiasm were less disrupted. Some changes precede developmental first eye opening.
Conclusions :
Our results highlight substantial abnormalities in RGC structure and function in NF1-OPG mice, and that early disruption in the retina’s “message” to the brain can precede anatomic defects. These findings may help explain the disparity seen between visual loss and tumor characteristics among NF1 patients. Ongoing experiments aim to more precisely identify the relationship of abnormal RGC activity with the several identified histologic abnormalities, providing additional targets for more effective treatments, and to determine developmental stages at which current promising treatments (e.g., MEK inhibitors) can be more effective. Ideally, these principles also will translate to earlier clinical detection of visual dysfunction.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.