Human deprivation amblyopia: treatment insights from animal models

Abstract

Amblyopia is a common visual impairment that develops during the early years of postnatal life. It emerges as a sequela to eye misalignment, an imbalanced refractive state, or obstruction to form vision. All of these conditions prevent normal vision and derail the typical development of neural connections within the visual system. Among the subtypes of amblyopia, the most debilitating and recalcitrant to treatment is deprivation amblyopia. Nevertheless, human studies focused on advancing the standard of care for amblyopia have largely avoided recruitment of patients with this rare but severe impairment subtype. In this review, we delineate characteristics of deprivation amblyopia and underscore the critical need for new and more effective therapy. Animal models offer a unique opportunity to address this unmet need by enabling the development of unconventional and potent amblyopia therapies that cannot be pioneered in humans. Insights derived from studies using animal models are discussed as potential therapeutic innovations for the remediation of deprivation amblyopia. Retinal inactivation is highlighted as an emerging therapy that exhibits efficacy against the effects of monocular deprivation at ages when conventional therapy is ineffective, and recovery occurs without apparent detriment to the treated eye.

Keywords: amblyopia; amblyopia therapies; animal models; monocular deprivation; neural plasticity (NP); neuropathology; recovery.

Figure 1

This graph plots the Snellen acuity achieved by the amblyopic eye as a function of the age at which a unilateral congenital cataract was removed and occlusion treatment began. Data demonstrate that, unlike other forms of amblyopia, effective treatment of deprivation amblyopia adheres to a short critical period in which, to promote optimal recovery from congenital MD, therapy must begin before the age of about 4 months (dashed vertical line). Therapy initiated beyond 4 months of age is associated with poor recovery outcomes. Graph displays results that were compiled by Birch and Stager (1988). Data originate from Beller et al., 1981 (triangles); Lewis et al., 1986 (squares); Pratt-Johnson and Tillson, 1981 (diamonds); Helveston et al., 1980 (hexagon); Awaya et al., 1979 (inverted triangles); Birch and Stager, 1988 (circles). CF indicates ability to count fingers; HM indicates the perception of hand movement.

Figure 2

Graphical representation of the surface area and tangential ocular dominance organization of V1 for human (A), macaque monkey (B), cat (C), rat (D), and mouse (E). Note the progressive decrease in overall surface area of V1 moving from human to mouse, and also the difference in organization observed across species. Scale bar is 10 mm. Images and organizational details of V1 originate from Adams et al. (2007) (human), Horton and Hocking (1996) (macaque monkey), Anderson et al. (1988) (cat), Duffy et al. (1998), Laing et al. (2015) (rat), and Airey et al. (2006) (mouse).

Figure 3

Ocular dominance histograms compare the effects of different rearing conditions that produce amblyopia. The ocular dominance of neurons sampled from the visual cortex of monkeys was measured by assessing the responsivity to stimulation of either the left or right eye. In normal animals there is a similar number of neurons connected to the right (group 1) and left (group 7) eye, with neurons connected to both eyes either equally (group 4) or somewhere in between (groups 2, 3, 5, 6; data from Kiorpes et al., 1998). Following monocular deprivation there is a strong shift in ocular dominance so that most neurons respond only to the non-deprived (fellow) eye (data from Hubel et al., 1977). Anisometropia produced by rearing with a contact lens placed in one eye also causes a shift in ocular dominance away from the affected eye, and like other amblyopia subtypes, reduces cortical binocularity (data from Kiorpes et al., 1998). Strabismus (esotropia or exotropia) does not result in disconnection of the affected eye but rather reduces the number of neurons responsive to both eyes – binocular cells (esotropia data from Kiorpes et al., 1998; exotropia data from Economides et al., 2021).

Figure 4

Data from cat and monkey revealing that the effect of intravitreal injection of TTX is reversible. VEPs (solid circles) measured from V1 in a cat (A; Duffy et al., 2023) and monkey (B; preliminary data) using scalp electrodes show a reduction to non-visual baseline levels (open circles) after TTX injection. Measurement of VEPs post-inactivation reveal a full recovery back to pre-inactivation levels for both species, indicating that the effect of inactivation on VEPs is temporary. OCT scans acquired from the monkey displayed in panel (B) demonstrate comparable retinal nerve fiber layer (RNFL) thickness between pre- and post-inactivation measurements (C). Similarly, individual registered b-scans suggest no change in retinal or optic nerve anatomy following injection (D). Monkey VEPs were collected for an experiment in which 4 TTX injections were delivered into the right eye over 4 weeks (one injection per week). Pre-inactivation VEPs were measured from right V1 at 8 months of age. Inactivation VEPs were measured 24 h after the first injection. Post-inactivation VEPs were taken 1 week after the final of four injections. OCT scans were acquired using the Spectralis OCT system (Heidelberg Engineering, Heidelberg, Germany), after pupils were dilated with 1% tropicamide. Scans acquired included a high resolution 55×45 degree raster scan, and 20×20 degree high speed raster scans centered on the optic nerve head and macula. Images were processed using neural network-based segmentation algorithms previously described (Srinivasan et al., 2022).