Caballero et al. [21] examined the effects of intraocular transplantation of human CD34+ cells in rodent models of ischemic retinal disease, including neonatal hyperoxia-induced retinopathy, retinal ischemia-reperfusion in adult mice (temporary central retinal artery occlusion induced by ophthalmic hypertension) and streptozocin-induced diabetes in adult rats. Transplanted cells co-localized with damaged host vasculature in all three models and vessel caliber and patency was improved in transplanted eyes compared to controls. Interestingly, there was no qualitative difference between intraocular and systemic route of administration, consistent with the ability of these cells to migrate to sites of tissue injury from peripheral blood.

Interestingly, BMSC transplantation may have applications in non-ischemic retinal disease. Otani et al. [22, 23] studied the effects of intravitreal injection of BMSCs in two models of inherited retinal degeneration, the rd1 mouse and the rd10 mouse. In both models, bone marrow-derived Lin− cells improved retinal thickness, retinal vascular density, and retinal function as measured by ERG. These findings suggest that transplantation of these cells may have effects not directly attributable to angiogenesis. In support of this hypothesis is the finding that BMSC-derived cells found at the site of RPE damage differentiate into RPE cells [14]. Genetic analysis of these cells suggest that this is not the result of fusion of bone marrow-derived cells with unlabeled, differentiated host cells.

If BMSCs capable of mitigating damage from eye injury already exist in the body normally, why would therapeutic transplantation be beneficial? One simple rationale is that the normal reparative response is limited by the number of available BMSCs. Since BMSCs can be expanded ex vivo, greater numbers of these cells may augment the normal host response [24]. A second rationale is that certain pathologic settings, such as diabetes, could affect the availability or ability of BMSCs to participate in restorative anigogenesis. Fewer HSCs are found in the peripheral blood of diabetic human subjects [25, 26] and in rats with experimental diabetes [27], suggesting that fewer of these stem cells may be available to participate in injury response in peripheral tissues. Hyperglycemia has been shown to impair survival and function of circulating blood-derived progenitor cells in vitro [28]. Further, transplantation of CD34+ cells isolated from diabetic subjects did not show beneficial effects seen with cells from nondiabetic subjects in rodent models of ischemic retinopathy [21].