A systematic review on transplantation studies of the retinal pigment epithelium in animal models

The retinal pigment epithelium (RPE) and the adjacent light-sensitive photoreceptors form a single functional unit lining the back of the eye. Both cell layers are essential for normal vision. RPE degeneration is usually followed by photoreceptor degeneration and vice versa. There are currently almost no effective therapies available for RPE disorders such as Stargardt disease, specific types of retinitis pigmentosa, and age-related macular degeneration. RPE replacement for these disorders, especially in later stages of the disease, may be one of the most promising future therapies. There is, however, no consensus regarding the optimal RPE source, delivery strategy, or the optimal experimental host in which to test RPE replacement therapy. Multiple RPE sources, delivery methods, and recipient animal models have been investigated, with variable results. So far, a systematic evaluation of the (variables influencing) efficacy of experimental RPE replacement parameters is lacking. Here we investigate the effect of RPE transplantation on vision and vision-based behavior in animal models of retinal degenerated diseases. In addition, we aim to explore the effect of RPE source used for transplantation, the method of intervention, and the animal model which is used. Methods: In this study, we systematically identified all publications concerning transplantation of RPE in experimental animal models targeting the improvement of vision (e.g., outcome measurements related to the morphology or function of the eye). A variety of characteristics, such as species, gender, and age of the animals but also cell type, number of cells, and other intervention characteristics were extracted from all studies. A risk of bias analysis was performed as well. Subsequently, all references describing one of the following outcomes were analyzed in depth in this systematic review: a-, b-, and c-wave amplitudes, vision-based, thickness analyses based on optical coherence tomography (OCT) data, and transplant survival based on scanning laser ophthalmoscopy (SLO) data. Meta-analyses were performed on the a- and b-wave amplitudes from electroretinography (ERG) data as well as data from vision-based behavioral assays. Results: original research articles met the inclusion criteria after two screening rounds. Overall, most studies were categorized as unclear regarding the risk of bias, because many experimental details were poorly reported. Twenty-three studies reporting one or more of the outcome measures of interest were eligible for either descriptive (thickness analyses based on OCT data; n = 2) or meta-analyses. RPE transplantation significantly increased ERG a-wave (Hedges’ g 1.181 (0.471–1.892), n = 6) and b-wave (Hedges’ g 1.734 (1.295–2.172), n = 42) amplitudes and improved vision-based behavior (Hedges’ g 1.018 (0.826–1.209), n = 96). Subgroup analyses revealed a significantly increased effect of the use of young and adolescent animals compared to adult animals. Moreover, transplanting more cells (in the range of 105 versus in the range of 104) resulted in a significantly increased effect on vision-based behavior as well. The origin of cells mattered as well. A significantly increased effect was found on vision-based behavior when using ARPE-19 and OpRegen® RPE. Conclusions: This systematic review shows that RPE transplantation in animal models for retinal degeneration significantly increases a- and b- wave amplitudes and improves vision-related behavior. These effects appear to be more pronounced in young animals, when the number of transplanted cells is larger and when ARPE-19 and OpRegen® RPE cells are used. We further emphasize that there is an urgent need for improving the reporting and methodological quality of animal experiments, to make such studies more comparable