Possible safety concerns from new stem cell, RPE vision loss study report

With stem cells for vision loss, first we want to be sure a treatment won’t make things worse.

Several teams around the globe are rigorously studying stem cell-based approaches to vision loss via regulatory-compliant studies including for macular degeneration with some results cautiously upbeat on safety from early phase analyses, but data from a new study on the use of human embryonic stem cell (hESC)-derived retinal pigmented epithelial cells (RPEs) for macular degeneration are concerning in some ways.

Figure 1, Mehat, et al. AAO Journal 2018; stem cells for vision
Figure 1, Mehat, et al. AAO Journal 2018

The paper in the AAO Journal from a team led by James W.B. Bainbridge at UCL is entitled “Transplantation of Human Embryonic Stem Cell-Derived Retinal Pigment Epithelial Cells in Macular Degeneration.”

The data here suggest that this kind of approach has some potential issues. While every stem cell study is unique, this paper also has important implications for the other studies out there using either hESC or human induced pluripotent stem cells (hIPSC) to make RPE for various forms of macular degeneration.

First, the bit of encouraging news was, “We found no evidence of uncontrolled proliferation or inflammatory responses,” which is something. You can see fundus images of the patients’ retinas in Figure 1.

However, they did not find a treatment-related improvement in vision so efficacy is not really supported so far with the analysis of the 12 patients for 13 weeks. Longer-term improvements are possible, but relatively less likely I think.

Also, there was a potential adverse event in a patient receiving the highest dose of RPEs. The authors wrote, “In one instance at the highest dose, localized retinal thinning and reduced sensitivity in the area of hyperpigmentation suggested the potential for harm.”

You can see supplemental Figure 3 below and note the drop in pigmented thickness in patient 10, especially as time went on in the study.

Supplemental Figure 3, Mehat, et al. AAO Journal 2018
Supplemental Figure 3, Mehat, et al. AAO Journal 2018

What does this mean? It sounds like the RPE at the higher dose may have actually made the retina less healthy and impaired vision, which is worrisome, but more data are needed to be sure.

Other events are described that potentially could be concerning too including some possible immune rejection in 2 patients and ectopic RPEs in a number of patient, although the latter apparently didn’t clearly affect function:

“A reduction in the pigmentation density at the site of transplantation, evident in 2 participants (at month 6 in patient 9 and at month 12 in patient 4) may reflect rejection of pigmented donor hESC-derived RPE cells, although no associated change in retinal function was apparent.”


“In our study, intraocular administration of carefully obtained hESC-derived RPE cells resulted in the development of pigmented foci in the vitreous cavity in 2 participants and on the surface of the inner retina in 2 participants, suggesting reflux of donor hESC-derived RPE cells from the subretinal compartment into the vitreous cavity. Despite the presence of preretinal or intravitreal pigmentation, no associated adverse effect was evident.”

I asked Dr. Bainbridge for his perspectives on these points and the paper overall, and he had these comments:

“The hyperpigmentation is consistent with the presence of donor cell material but we have no evidence that the cells are surviving.

We are not able to biopsy the engrafted area safely to determine donor cell survival with confidence.

We were not able to detect any improvement in retinal function in the hyperpigmented areas despite detailed perimetric analysis.

Further analysis will determine whether the intervention can protect sight in the longer term.

Localised retinal thinning and reduced function may be a consequence of the natural history or the intervention.

Detailed topographical analysis of structure and function is needed to determine the safety and potential efficacy of such interventions.”

The paper ends cautiously this way:

“The evidence of safety broadly supports the rationale for further studies to explore the impact of intervention at an earlier stage of degeneration when surviving photoreceptors cells may stand to benefit with improved function and survival. However, instances of focally reduced sensitivity and thinning in hyperpigmented retina at higher doses of hESC-derived RPE cells suggest the potential for harm and indicate that intervention at earlier stages of degeneration should be approached with caution.”

Although other similar studies have reported some adverse events (e.g. see here), they seemed mostly likely to be procedural rather than cell-related. The outcomes here warrant more thought and some extra caution moving forward.


  1. Paul, could you maybe talk about why you think they got this kind of result? Some other studies have done much better, so what was different here? Was their protocol different? Where did the cells come from? Was it something about their procedure?

    • I’m not sure why this paper seems different and more on the discouraging side. I need to study up on the nitty gritty of the different methods in the different papers. Maybe the starting cells were different, but I don’t recall. My impression was that this study maybe looked at retina in a different way too in following up. It had more patients that some of the others. I wish I knew what Bob Lanza’s program at Astellas was doing at this point too. If I have more insights on any of this moving forward I’ll let you know.

Leave a Reply