Mesenchymal stem cell therapy unable to rescue the vision from advanced Behcet's disease retinal vasculitis: report of three patients

Authors


Correspondence: Professor Fereydoun Davatchi, Rheumatology Research Center, Tehran University of Medical Sciences, Shariati Hospital, Kargar Avenue, Tehran 14114, Iran.

Email: fddh@davatchi.net

Abstract

Objective

Retinal vasculitis (RV) is the most aggressive lesion of ocular manifestations of Behcet's disease, seen in 32.1% of patients. Although visual acuity (VA) improves with early and aggressive treatment, in the long run it is seen in only 48% of patients. Mesenchymal stem cell (MSC) transplantation (MSCT) can theoretically reverse the RV process.

Patients and Methods

Three patients with advanced RV and very low VA were selected. Eyes selected for MSCT were legally blind (no useful vision) with severe retinal damage due to vasculitis, resistant to combinations of monthly pulse-cyclophosphamide (1000 mg) + azathioprine 2–3 mg/kg/day + prednisolone 0.5 mg/kg/day. After patient signed written consent, 30 mL of bone marrow were taken and cultured for MSC growth. After having enough MSCs in culture (4–5 weeks) and taking into consideration all safety measures, cells were injected in one eye of each patient (approximately 1.8 million MSCs). VA was measured. Disease Activity Index (DAI) was calculated for anterior uveitis (AU), posterior uveitis (PU) and RV.

Results

Visual acuity was light perception (LP) for two patients and finger count (FC) for the third. Follow-up at 1, 6 and 12 months were respectively LP/LP/FC at 0.5 m, no-light perception (NLP)/LP/LP, NLP/LP/NLP.

Discussion

Results showed a total failure of the procedure, essentially due to the late and advanced state of vasculitis. However, the autoimmune/inflammatory reaction was greatly controlled by the procedure.

Conclusion

Earlier cases have to be selected for further trials.

Introduction

Behcet's disease (BD) is a rare disease classified among the vasculitides. It is mainly seen along the Silk Road[1] Considering its rarity, the prevalence is rather high in Iran,[2] around 80 for 100 000 inhabitants.[3] Ocular manifestations are frequent, evident in 56.8% of patients. Retinal vasculitis (RV) is the most aggressive ocular lesion, seen in 32.1% of patients.[4] Although visual acuity (VA) improves with early and aggressive treatment (cytotoxics + corticosteroids), in the long run it effects 47.1% of men and 48.8% of women. In males, 5.6% of eyes lose their useful vision (legal blindness) and 8.6% become blind: a total of 14.2% of eyes. In females, 5.7% lose their useful vision and 6.2% become blind. RV is the most treatment-resistant ocular lesion. The inflammatory index of RV improved in 62% of males and 64.4% of females, by combination of cytotoxic agents and corticosteroids.[5] The remaining, 1/3 of patients had aggravated RV, leading to severe loss of vision or blindness. It is therefore necessary to seek other treatments for these resistant cases.

Mesenchymal stem cells (MSC) are able to differentiate to other cells. They can differentiate to osteoblasts, chondrocytes, and other cells of mesenchymal tissues.[6] They can immigrate to damaged tissues and there, under the influence of the local environment, differentiate to the specialized cells of the tissue and help the repair process.[7] MSCs have demonstrated the possibility of improving the autoimmune phenomenon, to reduce the inflammation, and to replace damaged cells. They are unable to stimulate the proliferation of lymphocytes and in vitro they inhibit up to 50% of that proliferation. They have therefore an immunoregulatory effect both in vitro and in vivo[8, 9] On the other hand, MSCs have been used in graft versus host disease (GVHD).[10-12] Studies show that MSCs have both an immunosuppressive and repairing effect. In 2002, Otani and colleagues demonstrated that bone marrow non-hematopoietic lineage cells (MSC) could help angiogenesis in deficient mice.[13] Kicic in 2003 showed that autologous CD90+ MSCs in the eyes of mice with retinal degeneration could change to photoreceptor cells.[14] In 2004, Smith[15] and also Otani and colleagues,[16] showed that these cells can rescue the retinal cone cells in retinitis pigmentosa of mice, a degenerative eye disease. Recently, the same has been done in adult humans, but with embryonic stem cells, giving encouraging results.[17] The importance of this work was to demonstrate the absence of side-effects, feared especially with embryonic stem cells than adult MSC. The MSCs can modulate and suppress autoimmune reaction by initiating the proliferation of immature antigen presenting cells (APC). These APCs will lead T cells toward anergy or regulatory (CD4+CD25+) cells.[18] MSCs have low immunogenicity due to the absence of human leukocyte antigen (HLA) class II antigens and co-stimulatory molecules (CD80, CD81).[19] They promote tissue repair by differentiating into the injured cell types, compensating their loss and secreting trophic factor.[20]

In our present study, autologous MSCs were separated from bone marrow of patients with BD having advanced retinal vasculitis, resistant to conventional cytotoxic combination therapy. As discussed by Smith,[15] the adult MSC derived from defective animals was effective in restoring the vasculature of the animal retina in retinitis pigmentosa and presumably the model could be applied to humans.

We present here the results of three cases of advanced BD RV treated with intravitreal injection of autologous MSCs.

Patients and Methods

Ethics and registration

The research carried out here with human subjects was in compliance with the Helsinki Declaration. It was approved by the Research Committee and the Ethical Committee of the Tehran University of Medical Sciences, and is registered under the ID 3086. It was then registered at ClinicalTrials.gov under the ID NCT00550498.

Patients

Two men (AA 34 and MBK 50 years old), and one woman (RK 50 years old) were selected for the study. The inclusion criteria were: having BD, RV resistant to the combination of monthly pulse cyclophosphamide with 1000 mg + daily azathioprine of 2–3 mg/kg body weight + daily prednisolone of 0.5 mg/kg body weight, being legally blind (no useful vision) because of ethical considerations and having a retinal or macular edema confirmed by fluorescein angiography and optical coherence tomography (OCT). The exclusion criteria were: no VA (complete blindness) and no retinal inflammation.

They were fully explained about the procedure, and after their signed written consent, they entered the study. The following parameters were calculated before the intervention and then regularly at each follow-up: VA on Snellen chart, disease activity index (DAI) of the anterior uveitis (AU), DAI of posterior uveitis (PU) and DAI of RV.[21] A total inflammatory activity index (TIAI) and a total adjusted disease activity index (TADAI) were also calculated. The formula to calculate the TIAI is right eye ([AU × 1] + [PU × 2] + [RV × 3]) + left eye ([AU × 1] + [PU × 2] + [RV × 3]). The formula to calculate TADAI is TIAI + right eye ([10 − VA] ×2) + left eye ([10 − VA] × 2).[22]

Sample collection and MSC expansion

Three to five weeks prior to injection, 30 mL of autologous bone marrow (BM) were obtained from each study patient. The BM mononuclear cells (MNC) were separated by Ficoll density gradient method. MNC were seeded in culture flasks with MSC medium, consisting of Dulbecco's modified Eagle's medium and 10% fetal bovine serum. Flasks were incubated at 37°C in a humidity chamber containing 5% CO2 and were fed by complete medium replacement every 4 days, until the confluence of fibroblast-like cells at the base of flasks. Thereafter the adherent cells were re-suspended using 0.025% trypsin and reseeded at 1 × 104 cells/mL. When cells reached confluence by the end of the second passage, they were incubated only with M199 medium for one more day. Cells were detached with trypsinization and washed with normal saline supplemented with 2% human serum albumin three times, then re-suspended at a density of 1–1.5 × 106 cells/mL.

Immunophenotyping

The expression of CD105, CD44, CD13 (MSC markers), CD34, CD45 (HSC markers), and CD31 (endothelial cell marker), were determined in culture-expanded MSCs using flow cytometry. Anti-CD44, CD45 and CD34 fluorescein isothiocyanate (FITC), anti-CD13 and CD31 phycoerythrin (PE) were all purchased from Dako (Glostrup, Denmark), along with anti-CD105, PE from Serotec (Milan, Italy). Flow cytometry was performed on a FacScan (Becton Dickenson, Franklin Lakes, NJ, USA). Data were analyzed with cellquest software (http://flowcytometry.med.ualberta.ca/PDF/FACScan%20%20Setup.pdf).

Safety assessment

Bacteriological tests were performed on samples after each passage, and before any injection (to make sure of -ontamination of samples). Before injection the viability of cells was assessed by methylene blue dye exclusion test.

Injection of MSCs

A mean volume of 0.3 mL containing approximately 1.8 million MSCs was injected intravitreously, in the selected eye of the patient, after local anesthesia. Patients were not hospitalized for the procedure, and went back home 1–4 h after the procedure.

Follow-up

The first checking after the procedure was 24 h, the second at 72 h (3 days), and the third at 1 week, for detection of a possible infection. Then the controls were checked every month for 1 year. Every month VA and different DAI were checked. At months 3, 6 and 12, fluorescein angiography, OCT and electroretinogram were performed.

Results

Before the intra-ocular MSC transplantation

Visual acuity for the site of transplantation was light perception (LP) for patient MBK (left eye), LP for patient AA (left eye), and finger count (FC) at 1.5 m for patient RK (right eye). The VA for the other eye was respectively, 3/10, FC at 3 m, and FC at 2 m.

Disease activity index of the anterior chamber for the site of transplantation was 3 for MBK, 0 (absence) for AA and RK. For the other eye, it was 0 for MBK, AA and RK.

Disease activity index of the posterior chamber for the site of transplantation was 3 for MBK, 0 for AA and 2 for RK. For the other eye, it was 0 for MBK, 1 for AA and 0 for RK.

Disease activity index of RV for the site of transplantation was 0 for MBK, 4 for AA and 9 for RK. For the other eye, it was 3 for MBK, 0 for AA and 10 for RK.

Total inflammatory activity index was 18 for MBK, 14 for AA and 61 for RK. TADAI was 52 for MBK, 53 for AA and 99 for RK. Figures 1-6 show the fundus photography before and after injection of fluorescein (fluorescein angiography) for MBK, AA and RK. Figures 7-9 show the electroretinogram for MBK, AA and RK.

Figure 1.

Fluorescein angiography before the injection: Patient MBK.

Figure 2.

Fluorescein angiography after the injection: Patient MBK.

Figure 3.

Fluorescein angiography before the injection: Patient AA.

Figure 4.

Fluorescein angiography after the injection: Patient AA.

Figure 5.

Fluorescein angiography before the injection: Patient RK.

Figure 6.

Fluorescein angiography after the injection: Patient RK.

Figure 7.

Electroretinogram before mesenchymal stem cell transplantation: Patient MBK.

Figure 8.

Electroretinogram before mesenchymal stem cell transplantation: Patient AA.

Figure 9.

Electroretinogram before mesenchymal stem cell transplantation: Patient RK.

Side effects and sequelae for transplanted eyes were for MBK: cataract, organization and hemorrhage in the vitreous, arterial and venous necrosis, atrophy of the retina, suspicion of retinal detachment, atrophy of the disc, and vascular necrosis of the periphery of the retina. For AA they were: cataract, arterial and venous necrosis of the retina, atrophy of the disc and scar in the macula. For RK they were: cataract, organization of the vitreous, atrophy of the disc and scar in the macula. For the other eye for MBK, they were: cataract, arterial necrosis of the retina, atrophy of the retina and disc, and peripheral vascular necrosis. For AA, they were organization of the vitreous, arterial and venous necrosis of the retina and atrophy of the retina. For RK, they were: posterior synechia, cataract, vitreous organization, arterial necrosis and atrophy of the retina.

One month after intra-ocular MSC transplantation

Visual acuity for the site of transplantation was LP for MBK (left eye), LP for AA (left eye) and FC at 0.5 m for RK (right eye). The VA for the other eye was respectively 7/10, FC at 4 m and FC at 1.5 m.

Disease activity index of the anterior chamber for the site of transplantation was 0 for MBK, AA and RK. For the other eye, it was also 0 for MBK, AA and RK.

Disease activity index of the posterior chamber for the site of transplantation was 1 for AA and 0 for RK. The posterior chamber was invisible for MBK. For the other eye, it was 0 for MBK, 1 for AA and 3 for RK.

Disease activity index of RV for the site of transplantation was 0 for AA, and 2 for RK. The retina was invisible for MBK. For the other eye, it was 3 for MBK, 1 for AA and 4 for RK.

Total inflammatory activity index was 7 for AA and 24 for RK. TADAI was 47 for AA and 50 for RK. TIAI and TADAI were not calculated for MBK because the posterior chamber and retina of the injected eye were invisible.

Side effects and sequelae for transplanted eyes were the same for MBK. For AA organization of the vitreous was added. For RK retinal detachment was added. For the other eye for MBK, they were the same. For AA, cataract and scar in the macula were added. For RK, they were the same.

Six months after intra-ocular MSC transplantation

Visual acuity for the site of transplantation was “no light perception” (NLP) for MBK (left eye), LP for AA (left eye) and for RK (right eye). The VA for the other eye was respectively 7/10, FC at 4 m and FC at 4.5 m.

Disease activity index of the anterior chamber for the site of transplantation was 0 for MBK, AA and RK. For the other eye, it was 0 for MBK, AA and RK.

Disease activity index of the posterior chamber for the site of transplantation was 1 for AA and 0 for RK. The posterior chamber was invisible for MBK. For the other eye, it was 0 for MBK, 1 for AA and 1 for RK.

Disease activity index of RV for the site of transplantation was 3 for AA and 0 for RK. The retina was invisible for MBK. For the other eye, it was 8 for MBK, 2 for AA and 11 for RK.

Total inflammatory activity index was 19 for AA and 35 for RK. TADAI was 58 for AA and 74 for RK. TIAI and TADAI were not calculated for MBK because the posterior chamber and retina of the injected eye were invisible.

Side effects and sequelae for transplanted eyes were for AA: peripheral vascular necrosis was added. For RK sequelae were the same. The posterior chamber was invisible for MBK. For the other eye for MBK, they were the same. For AA, peripheral vascular necrosis was added. For RK, retinal atrophy was added.

One year after intra ocular MSC transplantation

Visual acuity for the site of transplantation was NLP for MBK (left eye), LP for AA (left eye) and NLP for RK (right eye). The VA for the other eye was respectively 7/10, FC at 2.5 m and FC at 2 m.

Disease activity index of the anterior chamber for the site of transplantation was 0 for MBK, AA and RK. For the other eye, it was 0 for MBK, AA and RK.

Disease activity index of the posterior chamber for the site of transplantation was 0 for MBK, 2 for AA and zero for RK. For the other eye, it was 0 for MBK, 1 for AA and 0 for RK.

Disease activity index of RV for the site of transplantation was 0 for MBK, AA and RK. For the other eye, it was 0 for MBK and AA, and 5 for RK.

Total inflammatory activity index was 0 for MBK, 6 for AA and 15 for RK. TADAI was 26 for MBK, 46 for AA and 55 for RK.

Side effects and sequelae for transplanted eyes were the same for MBK, AA and RK. It was same for the other eye of MBK and AA. For RK, detachment was added.

Discussion

Our experience of MSCs injected locally in the posterior chamber (uvea) was a total failure, regarding VA. Contrary to our expectations, the injected cells did not build new vessels and did not transform to cone cells. They inhibited the inflammation of uvea in MBK's injected eye, which did not recur in the follow-up, up to 1 year. It also controlled the macular edema of the injected eye. A periphlebitis appeared at 1 month, but disappeared later (control at 2 months). The good result on the inflammation is to be attributed to the MSCs, but it may also be attributed to the treatment by cytotoxics and steroids. However, the good eye (VA 7/10), the non-injected eye, had a flare of RV at 6 months, while the injected eye remained non-inflammatory. AA had a peripheral retinitis which disappeared at the 1 month control and did not recur. RK had 2+ cells in the vitreous before MSC injection along with periarteritis and periphlebitis of the retina that disappeared after MSC and did not recur up to 1 year of follow-up. The non-injected eye had the same inflammation that continued during the follow-up. These results show the capacity of immune modulation of the MSCs. The control of inflammation was not due to the cytotoxic and steroid treatment, because the inflammation of non-injected eyes continued.

Animal experiments on experimental autoimmune encephalomyelitis (animal model of multiple sclerosis) demonstrated that the best results were obtained at the beginning of the disease, and then at the peak of attacks,[23] which can be interpreted as essentially a preventive action of the procedure rather than curative action.

Our cases had very advanced eye lesions, with many sequelae. If the above finding can be true in humans, then our cases were not good cases, because they were selected too late for MSCs to be helpful. In the future, earlier cases, with more active lesions and fewer sequelae, have to be selected.

As these patients were the first human subjects to receive MSCs as local injections for BC RV, we are still very far from being able to find the right dose (total number of cells injected) of MSCs to inject. Another limitation factor is the volume of the suspension. In a human study, Schwartz et al.[17] used 5 × 104 cells in a volume of 150 μL, injected in the subretinal space. We injected here 10 times more cells, in twice the volume, but intra-vitreously. Further studies have to find the optimal number of cells to be injected, and the optimal volume of cell suspension.

A last concern remains, raised by one of the authors: the discovery of retinal detachment in two of the patients, MBK and RK. Was the retinal detachment due to the MSCs, multiplying exaggeratedly under the retina, and causing subsequently its detachment? This side effect has not been observed in animal experiments. It has not also been observed in human experiments with the embryonic stem cells,[17] which have higher tendencies to multiply. In these patients, the cells were injected under the retina. If the proliferation of cells could produce a detachment, these patients who had the cells directly injected under the retinal membrane would have been more prone to the detachment. On the other hand, we did a check on 100 consecutive BD patients with ocular lesions. Among those with advanced eye lesions, 14% had retinal detachment. In blind patients the percentage increased to 50%. It is also important to notice that RK had retinal detachment in the non-injected eye too, showing that retinal detachment was not due to the local proliferation of injected MSCs.

Conclusion

Mesenchymal stem cell transplantation may help patients with BD and resistant ocular manifestations from combinations of cytotoxic drugs and steroids or biologic agents. Proceeding trials have to select patients at earlier stages (more inflammation and fewer sequelae).

Acknowledgement

The authors thank Sina Cell for providing the facilities for MSC culture.

Ancillary