Postmortem Retinal Structural and Metabolic Analysis After Human Embryonic Stem Cell–derived Retinal Pigment Epithelium Transplantation in a Patient With Stargardt Disease

Abstract

Purpose: Stargardt disease is an inherited form of retinal degeneration characterized by early-onset central vision loss. This report describes the long-term retinal histologic characteristics of a patient with Stargardt disease after human embryonic stem cell–derived retinal pigment epithelium (hESC-RPE) transplantation. Methods: Eyes were obtained postmortem from an 80-year-old legally blind male patient with Stargardt disease. The patient had previously undergone hESC-RPE transplantation of his left eye, and the right eye (untransplanted) served as a control. Fluorescence lifetime imaging microscopy was used to study retinal structure and metabolic activity. Staining of the retina for lipofuscin was performed using the Armed Forces Institute of Pathology method. Results: Metabolic analysis showed that the macular area had more oxidative phosphorylation relative to the mid-peripheral retina, in both the transplanted and untransplanted eye. Melanin-laden RPE cells were detected in the transplanted eye, which correlated with a pigment band present on fundus images post–hESC-RPE transplantation. Conclusions: This case description illustrates the long-term survival of subretinal hESC-RPE cells post-transplantation in a patient with Stargardt disease.

Introduction

Stargardt disease is one of the most common forms of inherited retinal degeneration, resulting in early-onset macular degeneration, and is characterized by central vision loss. Although there is no effective treatment for Stargardt disease at present, several approaches are being tested, including stem cell transplantation. Several stem cell transplantation clinical trials for Stargardt disease have taken place, in which a variety of approaches have been used, including implantation of human embryonic stem cell–derived retinal pigment epithelium (hESC-RPE),^1–7 implantation of autologous bone marrow–derived stem cells, ⁸ and implantation of adipose tissue–derived mesenchymal stem cells. ⁹ Moreover, histologic analyses to study stem cell survival have been carried out in eyes with age-related macular degeneration,^10,11 but studies of this kind have not been reported in Stargardt disease. In this case report we present the long-term histologic findings in postmortem eyes from a patient with Stargardt disease who had undergone hESC-RPE transplantation.

Case Report

The patient was an 80-year-old man with a medical history of Stargardt disease, Parkinson’s disease, and Lewy body dementia. He was clinically diagnosed with Stargardt disease at the age of 18 years, Parkinson’s disease at the age of 70 years, and Lewy body dementia at the age of 73 years. He was declared legally blind in his late 20s and his vision continued to deteriorate throughout his life, which led to his retirement from work at the age of 64 years. Genetic testing at the age of 69 years showed a pathogenic heterozygous mutation (G1961E) in the ABCA4 gene. The patient had a heterozygous G>A nucleotide substitution at exon 42 in which glycine codon (GGA) was substituted with glutamic acid (GAA) at amino acid position 1961, or c.5882 G>A at the complementary DNA level, or pGly1961Glu or G1961E at the protein level. No second mutation was identified. One of his 2 sisters had the same mutation.

The patient underwent hESC-RPE implantation in his left eye at the age of 71 years as part of a phase I/II clinical trial. The transplanted hESC-RPE cells were derived from the MA09-hRPE cell line.^1,2 Ocular examination before hESC-RPE transplantation revealed counting fingers at 1 ft in his right eye and at 2.5 ft in his left eye, intraocular pressure of 13 mm Hg in both eyes, 1+ nuclear sclerotic cataracts bilaterally, cup-to-disc ratio of 0.3, and geographic atrophy in the macula surrounded by pisciform flecks bilaterally. Additionally, the right eye (untransplanted) showed RPE hyperplasia in the center of the macula. Post–hESC-RPE transplantation, there were no reported serious adverse events attributable to the transplanted cells. ¹ Despite having undergone transplantation in his left eye, the patient’s vision failed to improve. He also underwent cataract extraction and lens implantation in both eyes 2 years after the hESC-RPE implantation, and YAG capsulotomy in the left eye the following year. The patient died at the age of 80 years (see Supplementary Figure 1 for timeline).

After receiving consent from the family, the patient’s postmortem, formalin-fixed eyes were assessed for histologic characteristics. Gross histologic examination showed the presence of atrophy in the macular region in both eyes (Figure 1A and 1C). The right eye (untransplanted) showed pigmentation at the center of the atrophy. The left eye (transplanted) showed pigmentation in the hESC-RPE–transplanted region. These findings correlated with those observed on the last fundus images of the right and left eye obtained at 2 years post-transplantation (Figure 1B and 1D) and with the changes seen on fundus images from pretransplantation to 6 months and 15 months post-transplantation in the left eye (Figure 1E, 1F, and 1G).

Figure 1.

The case was a patient with clinically diagnosed Stargardt disease who underwent human embryonic stem cell–derived retinal pigment epithelium (hESC-RPE) transplantation in the left eye at age 71 years (right eye remained untransplanted), and died at age 80 years. Top, Images used for gross histologic examination of the postmortem, formalin-fixed (A) right eye and (C) left eye, and the last fundus photographs of the (B) right eye and (D) left eye, obtained from the patient 2 years post-transplantation at age 73 years. Arrows indicate the area of RPE hyperplasia. Arrowheads point to pigmentation at the area of hESC-RPE transplantation. Middle, Fundus photographs of the patient’s left eye (E) before hESC-RPE transplantation (baseline), (F) 6 months post-transplantation, and (G) 15 months post-transplantation. Note the absence of pigmentation in the baseline fundus image. The broken black circle indicates the region of hESC-RPE transplantation. The broken white arrows show the region of atrophy. Bottom, Corresponding optical coherence tomography images, with boxed areas indicating the region of interest. The fundus images in E, F, and G were acquired, with permission, from the authors of Schwartz et al. ¹

Histologic analysis was performed using fluorescence lifetime imaging microscopy (FLIM) and light microscopy. The FLIM settings and the phasor signals for melanin and lipofuscin were determined according to those used in previous studies.^12–14 In both the transplanted eye and the untransplanted eye, light microscopy with hematoxylin and eosin staining revealed atrophy of the inner and outer photoreceptor segments, the outer nuclear layer, and the outer plexiform layer, with scanty RPE cells in the macular area. RPE was absent in the atrophic part of the macula in both eyes, but was present in the hyperpigmented areas (superior macula in the transplanted eye and the center of the macula in the untransplanted eye).

In the mid-peripheral retina in both eyes, all retinal layers had cuboidal RPE and light brown granules within the RPE cells. These granules correspond to lipofuscin deposits, as evident on the FLIM images, which show the signal for N-retinylidene-N-retinylethanolamine (A2E) (a component of lipofuscin) within the mid-peripheral RPE (Figure 2A and 2B).

Figure 2.

Mid-peripheral retina of the patient’s untransplanted right eye. (A) Hematoxylin and eosin staining shows cuboidal retinal pigment epithelium (RPE) (black arrows) and all the retinal layers. The asterisk indicates artifact space. (B) Fluorescence lifetime imaging microscopy shows an orange signal from the RPE, which corresponds to the signal for N-retinylidene-N-retinylethanolamine (A2E; a component of lipofuscin) (yellow arrowheads) in the same area of the retina as shown in A. Top left inset in (B), Phasor representation of lipofuscin (area denoted by circle). Abbreviations: INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer; ONL, outer nuclear layer; OPL, outer plexiform layer; PR, photoreceptor layer; RNFL, retinal nerve fiber layer.

Additionally, staining of the cells for lipofuscin using the Armed Forces Institute of Pathology (AFIP) stain revealed that lipofuscin was present in the mid-peripheral RPE cells; however, no AFIP staining for lipofuscin was evident in the RPE of the macular area (Figure 3A and 3B). The transplanted area of the left eye (the superior aspect of the macula) showed flat RPE cells with dark granules arranged as a monolayer in some areas and as a multilayer in other areas. These dark brown RPE granules correspond to melanin, as shown by FLIM imaging (Figure 4A, 4B, and 4C).

Figure 3.

Mid-peripheral retina of the patient’s left eye after human embryonic stem cell–derived retinal pigment epithelium (hESC-RPE) transplantation. Armed Forces Institute of Pathology staining was performed for detection of lipofuscin. (A) Pink staining of the mid-peripheral RPE in the transplanted eye indicates the presence of lipofuscin (arrows). The asterisk indicates artifact space. (B) Lack of pink staining in the RPE of the macular area in the transplanted eye indicates the absence of lipofuscin (arrowheads). Abbreviations: INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer; ONL, outer nuclear layer; OPL, outer plexiform layer; PR, photoreceptor layer; RNFL, retinal nerve fiber layer.

Figure 4.

Macular region of the patient’s left eye after human embryonic stem cell–derived retinal pigment epithelium (hESC-RPE) transplantation compared with the untransplanted right eye. (A) Hematoxylin and eosin (H&E) staining of the left eye shows pigmented RPE (arrows) and outer retinal atrophy (asterisk). Given that this area was atrophic and became hyperpigmented after transplantation, it can be assumed that these cells are proliferated transplanted cells. Note the irregular pattern of cells with monolayer and multilayer presentations. Top left inset in (A), Fundus image of the transplanted eye. The white line shows the cross-section obtained for histologic analysis. The white box shows the specific location of the histologic sections shown in A and B. (B) Fluorescence lifetime imaging microscopy imaging of the left eye shows the signal for melanin (pink) in the RPE (yellow arrowheads) and choroid (white arrowheads). Top left inset in (B), Phasor representation of melanin (area denoted by circle). There are 2 foci on the phasor plot, one arising from the presence of bound reduced nicotinamide adenine dinucleotide, or NAD(P)H (orange arrow), and the other from melanin (inside the circle). The position of melanin on the phasor plot was determined from our previous work. No foci for the lipofuscin signal N-retinylidene-N-retinylethanolamine (A2E) could be detected on the phasor plot, suggesting a lack of its expression in these RPE cells. (C) H&E staining of the corresponding macular area in the patient’s untransplanted right eye. Arrows point to the area of atrophy; note that RPE is absent. Top left inset in (C), Fundus image of the right eye. The white line shows the cross-section obtained for histologic analysis. The white box shows the specific location of the histologic section. Abbreviations: INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer; ONL, outer nuclear layer; OPL, outer plexiform layer; PR, photoreceptor layer; RNFL, retinal nerve fiber layer.

To analyze the retinal metabolic state, FLIM imaging was used to determine the levels of bound reduced nicotinamide adenine dinucleotide, or NAD(P)H (a metabolic parameter), in both the transplanted eye and the untransplanted eye, comparing 8 regions in the mid-peripheral retina with 8 regions in the macula (4 on each side of the fovea) (n = 8). The width of each region was kept constant at 150 µm. The FLIM settings for metabolic imaging were based on those reported in our previous publication. ¹² Two-sample t-tests were used for statistical analysis.

In the transplanted eye, the mean ± SD percentage of bound reduced NAD(P)H was 58.4 ± 2.3% in the macula and 54.1 ± 2.3% in the mid-peripheral retina (P = .002). In the untransplanted eye, the mean ± SD percentage of bound reduced NAD(P)H was 61.7 ± 1.1% in the macula and 57.4 ± 3.2% in the mid-peripheral retina (P = .003). This indicates that there was a statistically significant difference in the metabolic state between the atrophic macula and the mid-peripheral retina.

In comparing the inner retina (inner nuclear layer, inner plexiform layer, ganglion cell layer, and retinal nerve fiber layer) of the mid-peripheral retina with the macular area, the mean ± SD percentage of bound reduced NAD(P)H was 54.3 ± 1.9% in the inner mid-peripheral retina and 58.4 ± 2.3% in the macular area of the transplanted eye. The mean ± SD percentage of bound reduced NAD(P)H was 59.3 ± 2.6% in the inner mid-peripheral retina and 61.7 ± 1.1% in the macular area of the untransplanted eye. These findings show that there was a statistically significant difference in the metabolic state in these regions, both in the transplanted eye (P = .001) and in the untransplanted eye (P = .03).

Conclusions

This patient with Stargardt disease underwent hESC-RPE transplantation as part of a phase I/II clinical trial. Following the clinical trial, the fundus images demonstrated an area of pigmentation in the transplanted region of the left eye that was not present before transplantation. ¹ Our goal in this study was to assess the long-term survival of hESC-RPE cells in the transplanted eye using histology. In comparing the gross histopathologic characteristics of the transplanted eye with the last available fundus images of the transplanted eye, both showed hyperpigmentation in the transplanted area, which was used as a guide for sectioning and imaging the transplanted area. Given the absence of hyperpigmentation before transplantation, and pigment formation in the atrophic macula at the hESC-RPE–transplanted site, it can be assumed that these findings were indicative of the presence of proliferated hESC-RPE cells post-transplantation. In the untransplanted eye, however, macular hyperpigmentation was present in the fundus images both before and after hESC-RPE transplantation, which correlates with the characteristics of RPE hyperplasia.

A histologic study of the hESC-RPE–transplanted eye and untransplanted eye was performed. Previous histologic studies have detected the presence of lipofuscin in RPE cells of patients with Stargardt disease and fundus flavimaculatus, detected using special stains and autofluorescence.^15–21 In our study, we used FLIM to detect lipofuscin and melanin. The findings from FLIM imaging showed that the A2E signal for lipofuscin, but not the signal for melanin, was detected in the mid-peripheral RPE of both the transplanted eye and the untransplanted eye. A2E is a component of lipofuscin and is one of the most predominant components of lipofuscin in RPE. ²² A previous study determined the phasor position of A2E by using synthetic A2E. ¹³ The phasor signal observed in that study correlated with the phasor signal detected in our study, thereby confirming the presence of lipofuscin within the mid-peripheral RPE. We additionally used the AFIP method of staining for lipofuscin, which further reinforced our findings. In the hyperpigmented areas of the macula, melanin signal was seen in both the transplanted eye and the untransplanted eye, but the A2E signal was undetected in these areas, indicating that lipofuscin was predominant in the mid-peripheral RPE and melanin was predominant in the macular RPE.

The FLIM technique can be used to assess the retinal metabolic state by detecting molecules such as NAD(P)H. Several studies using FLIM have revealed whether the cells predominantly use glycolysis or oxidative phosphorylation as the mechanism for metabolic activity. A higher percentage of bound (versus unbound) NAD(P)H indicates a predominance of oxidative phosphorylation, whereas a lower percentage suggests a predominance of glycolysis.^12,23–27 In our previous mouse studies using normal (wild-type) mice and mice with retinal degeneration (rd10), a comparison of the peripheral retina and the posterior retina did not demonstrate a significant difference in metabolic activity.^12,27 However, in this study, we determined that in both eyes of the patient with Stargardt disease, the macula showed more oxidative phosphorylation when compared to the mid-peripheral retina. Several previous studies in patients with Stargardt disease investigated the presence of retinal flecks using fluorescence lifetime imaging ophthalmoscopy.^28–30 However, glycolysis and oxidative phosphorylation were not mentioned in those studies. Our study compared the retinal metabolic state in both the transplanted eye and the untransplanted eye. In comparing the transplanted and untransplanted eye, oxidative phosphorylation was found to be higher in the untransplanted eye, both in the macula and in the mid-peripheral retina. It was unclear whether this could be attributed solely to the effects of transplantation of hESC-RPE cells in the transplanted eye, since we observed the change in the mid-peripheral retina as well.

Several strategies were previously used to differentiate the donor cells from recipient cells; one of the most common ways was fluorescence in situ hybridization (FISH) of the sex chromosomes if the donor and the recipient belonged to different sexes.^11,31,32 A limitation of our study is that we could not potentially differentiate the female donor cells from the recipient RPE cells using FISH; therefore, we could not rule out the possibility that reactive hyperplasia of the host RPE cells was the source of pigmented cells in the transplanted eye. Another limitation is that the postmortem eyes of the patient discussed herein were already fixed in formalin, and FLIM images were obtained after deparaffinization of slides. In the future, potential imaging with fresh eyes will be a better predictor of metabolic activity.

The characteristics described in this case report of a patient with Stargardt disease reveal that RPE cells were present at the site of hESC-RPE transplantation 9 years after the procedure. In addition, findings from the metabolic analysis reveal a relatively higher level of oxidative phosphorylation in the macular area when compared to the mid-peripheral retina in both the transplanted eye and the untransplanted eye of this patient.