Evaluating the Impact of Mesenchymal Stem Cell Therapy on Visual Acuity and Retinal Nerve Fiber Layer Thickness in Optic Neuropathy Patients: A Comprehensive Systematic Review and Meta-Analysis

Major Points and Findings:

According to the authors, this paper is the first systematic review and meta-analysis of stem cell therapy in people with optic neuropathy (optic nerve atrophy of any cause). It pools every published human report the authors could find and addresses two questions. The first is whether visual acuity changed. The second is whether the retinal nerve fibre layer, the layer of nerve fibres that forms the optic nerve, became thicker.

Average visual acuity improved by about 0.25 logMAR (roughly two and a half chart lines) while the nerve fibre layer did not change. The evidence behind this result is weak. There are fewer than 100 treated eyes and no randomised trial, and the paper’s own sensitivity analysis shows the acuity result is no longer statistically significant if one study is removed. The paper also contains internal inconsistencies, listed below.

Aim:

To estimate, by meta-analysis of published reports, the change in visual acuity (logMAR) and retinal nerve fibre layer (RNFL) thickness after stem cell therapy in optic neuropathy.

Methods:

The authors, from Iran University of Medical Sciences, followed the PRISMA guideline and searched PubMed, Scopus and Google Scholar from inception to June 2024.

Reports were included if patients had a diagnosis of optic neuropathy, eyes were treated with stem cells, visual acuity or RNFL thickness was reported, and the paper was published in English. Two researchers extracted data independently.

Quality was assessed with the Mixed Methods Appraisal Tool. The authors rated the included papers as moderate quality, noting that basic facts such as the number of cells injected were often not stated.

The effect size was the mean change from baseline. Each eye counted as one sample, with generalised estimating equations to allow for eyes within a study being related. The abstract names STATA 16 as the software. The methods name SPSS 24. There was no comparison against untreated eyes.

The paper’s table lists eight reports published between 2015 and 2023:

• Three reports from the SCOTS group in the United States (Weiss and colleagues), namely two single-patient case reports and a series of 10 patients (20 eyes) with ischaemic optic neuropathy. All used the patient’s own bone marrow concentrate by combinations of retrobulbar, subtenon, intravitreal, intra-optic nerve and intravenous routes, without reported cell markers.
• Oner and colleagues, 2019: 4 eyes of 4 patients, adipose-derived MSCs placed in the suprachoroidal space.
• Limoli and colleagues, 2021: 35 eyes of 25 patients with progressive glaucomatous optic neuropathy, adipose stromal vascular fraction, suprachoroidal.
• Kahraman and colleagues, 2020: 29 eyes of 23 patients, umbilical cord-derived MSCs, suprachoroidal. Causes were mixed, comprising intracranial hypertension (6), glaucoma (5), traumatic optic neuropathy (3), optic atrophy after an intracranial mass (3), methanol poisoning (2), diabetic optic atrophy (2), retinal vein occlusion (1) and ischaemic optic neuropathy (1). This was the only report using umbilical cord cells.
• Vilela and colleagues, 2021: 2 patients with advanced open-angle glaucoma, own bone marrow MSCs, intravitreal.
• Pastor and colleagues, 2023: 5 patients with acute ischaemic optic neuropathy, donor bone marrow MSCs, intravitreal.

Every study delivered cells into or next to the eye. None tested intravenous or intrathecal delivery alone.

Results:

Pooled mean age was 53.2 years (five papers) and about 61% were male. Six papers reported acuity and four reported RNFL thickness. Follow-up is described as one year, or 15 months in one study.

Visual acuity: Mean logMAR went from 0.90 (SD 1.23) before treatment to 0.65 (SD 0.78) after, a change of 0.25, with p below 0.001. Lower logMAR means better vision. Vision improved in every study except two. Vilela and colleagues found no improvement, and the results section states that the umbilical cord MSC study by Kahraman and colleagues showed a worsening of acuity.

Nerve fibre layer: Mean RNFL thickness went from 77.56 µm (SD 23.02) to 82.55 µm (SD 15.46), about 5 µm, with p = 0.240. The change was not statistically significant.

Neither age (p = 0.991) nor sex (p = 0.345) was related to the change in acuity, and neither was related to the change in RNFL thickness.

Sensitivity analysis: Removing studies one at a time did not change the result, with one exception. When the 2017 SCOTS series by Weiss and colleagues was left out, the change in visual acuity was no longer significant. The pooled result therefore rests on one uncontrolled series that used mixed delivery routes.

Safety: The discussion states that no adverse events, rejection or abnormal cell growth were reported in the included studies. Details elsewhere in the same paper contradict this. One patient in the Vilela study was withdrawn because of retinal detachment with proliferative vitreoretinopathy, and the included Pastor trial reported an epiretinal membrane that led to retinal detachment in 1 of 5 patients.

The paper contains the following inconsistencies:

• The abstract reports 66 eyes in seven papers. The results text says 62 eyes were analysed and that 94 eyes of 66 patients were treated. The table lists eight reports, where the abstract counts seven.
• The abstract gives p = 0.174 for RNFL. The results table gives 0.240 (0.174 is the p-value for the age analysis).
• The discussion gives the acuity change as 0.88 to 0.64. The results give 0.90 to 0.65.
• The results say acuity worsened in the Kahraman study. The discussion says the same study found improved acuity and visual fields with no RNFL change.

This summary follows the results section and tables.

Conclusions:

The authors conclude that stem cell therapy “may improve” visual acuity in optic neuropathy. They add that reaching phase III trials will take considerable effort, that unified dosing criteria and strict guidelines are needed first, and that MSC-derived exosomes should be tested.

The authors list these limitations. There were few studies, with small samples. Many participants were legally blind, and acuity is hard to measure reliably in such patients, which raises the risk of bias. Cell types, doses and routes were not compared. The OCT device was unknown in most studies. Visual fields, ERG and eye pressure were not analysed. No long-term safety data exist, and immune reactions and tumour formation are named as open concerns.

Background Information:

Acuity improved without any thickening of the nerve fibre layer. If cells were regrowing optic nerve fibres, the RNFL should thicken, and it did not. The authors say it is unclear whether acuity gains come from the cells directly. They describe a paracrine mechanism in which MSCs release growth factors, anti-inflammatory signals and vesicles that may help surviving nerve cells work better. They note that MSCs have low rates of migration and differentiation.

On delivery into the eye, the authors comment that intravitreal injection is simple but poorly suited to cells, because the vitreous must stay clear, and suggest the MSC secretome may avoid such complications.