Article Text

Download PDFPDF

Synthetic biodegradable alternatives to the use of the amniotic membrane for corneal regeneration: assessment of local and systemic toxicity in rabbits
  1. Charanya Ramachandran1,
  2. Virender S Sangwan1,
  3. Ilida Ortega2,
  4. Upendra Bhatnagar3,
  5. Sadik Mohmad Abdulhamid Mulla3,
  6. Rob McKean4,
  7. Sheila MacNeil5
  1. 1 Sudhakar and Sreekanth Ravi Stem Cell Laboratory, LV Prasad Eye Institute, Kallam Anji Reddy campus, LV Prasad Marg, Hyderabad, Telangana, India
  2. 2 School of Clinical Dentistry, Claremont Crescent, University of Sheffield, Sheffield, UK
  3. 3 Preclinical Division, Vimta Labs Ltd, Hyderabad, Telangana, India
  4. 4 The Electrospinning Company Ltd, Rutherford Appleton Laboratory, Oxford, UK
  5. 5 Department of Materials Science and Engineering, Kroto Research Institute, North Campus, University of Sheffield, Sheffield, UK
  1. Correspondence to Dr Sheila MacNeil, Department of Materials Science and Engineering, Kroto Research Institute, North Campus, University of Sheffield, Sheffield S3 7HQ, UK; s.macneil{at}


Aim The aim of this study was to assess the local and systemic response to poly-lactic co-glycolic acid (PLGA) 50:50 membranes, developed as synthetic biodegradable alternatives to the use of human donor amniotic membrane in the treatment of limbal stem cell deficiency.

Methods PLGA membranes of 2  cm diameter and 50  µm thickness were placed on one eye of rabbits and secured in place using fibrin glue and a bandage contact lens, suturing the eye close with a single stitch. Control animals were treated identically, with the absence of the membranes. Plain and microfabricated electrospun membranes (containing micropockets which roughly emulate the native limbal niche) were examined over 29 days. All animals were subjected to a detailed gross and histopathological observation as well as a detailed examination of the eye.

Results Application of the membranes both with and without microfabricated pockets did not adversely affect animal welfare. There was complete degradation of the membranes by day 29. The membranes did not induce any significant local or systemic toxicity. Conjunctival congestion and corneal vascularisation were noted in a few control and PLGA-treated animals. Intraocular pressure was normal and the retinal status was unaltered. The ocular surface was clear and intact in all animals by the end of 29  days.

Conclusion Membranes of 50:50 PLGA can be safely applied to rabbit corneas without inducing any local or systemic toxicity and these break down completely within 29 days.

  • cornea
  • ocular surface
  • stem cells

Statistics from


Limbal stem cell deficiency (LSCD) can be caused by injuries, surgery, disease and developmental anomalies causing loss of corneal epithelial stem cells located in the limbus and leading to conjunctivalisation of the ocular surface.

Treatments for LSCD range from grafting limbal tissues to transplanting cultured limbal epithelial cells (CLET) to more recently transplanting small pieces of limbal explants directly onto the ocular surface (simple limbal epithelial transplantation (SLET)).1–3 SLET unlike CLET obviates the need for clean room facilities and cell culture technicians, therefore becoming much more accessible.4 Both CLET and SLET procedures are successful in around 70% of patients with LSCD from chemical burns.5–7

Human amniotic membrane (hAM) is commonly used in CLET and SLET as a sacrificial substrate. It provides support to the transplanted cells ensuring good attachment and cell survival on the ocular wound bed. SLET has radically simplified the treatment of LSCD, but further improvements could widen availability. A synthetic membrane would be more accessible and safer than using hAM.

Alternative carrier materials have been studied for the transplantation of limbal stem cells including coated contact lenses, compressed collagen and poly-lactic co-glycolic acid (PLGA) membranes.8 9 PLGA membrane has been extensively used in the clinic as dissolvable sutures.8 Our previous work shows that electrospun PLGA membranes based on a 50:50 ratio of glycolic acid to lactic acid broke down within 4–6 weeks in vitro, supported limbal cell culture and expansion as well as9 hAM in terms of stem cell expansion and maintenance without the need for feeder cells.9 10 We also showed transfer of cells from these membranes to cornea organ culture models in vitro.

Limbal stem cells are believed to be located within crypts in the limbus and in contact with supporting cells which help them retain their stemness.11–13 To mimic the native tissue, we designed a membrane incorporating microfabricated pockets to provide protection to the limbal explants that are transplanted onto the ocular surface.14

In this study, we applied PLGA membranes both with and without microfabricated pockets to the corneas of rabbits for 29 days. Experiments were done in collaboration with an accredited contract research organisation, Vimta Labs Ltd, following Schedule ‘Y’, Drugs and Cosmetic (IIAmendment) Rules, for assessing topical and systemic toxicity.

Both membranes broke down completely within 29 days without adverse effects.


PLGA membranes were produced by The Electrospinning Company Ltd, UK. PLGA 50:50 (Corbion, the Netherlands) of molecular weight 44 kg/mol was used in this study. Membranes were gamma irradiated (Steris AST, Swindon, UK), packaged and sent to India by World Courier (London, UK).

New Zealand rabbits were sourced from Delve Labs, Hyderabad.

Bandage contact lens and Ethicon non-absorbable surgical were from Johnson and Johnson Vision care, USA, and Tisseel kit (fibrin glue) from Baxter, Austria.


The study was performed in compliance with Organisation for Economic Co-operation and Development Principles of Good Laboratory Practice for the testing of chemicals as specified by International (C(97)186/Final) Legislation. The study protocol was abided by the recommendations of the Committee for the Purpose of Control and Supervision of Experiments on Animals guidelines.

Experimental design

Rabbits (age 10–13 weeks) were acclimatised for 5 days prior to surgery then randomised based on their weight. Fundus examination, intraocular pressure (IOP) measurement and fluorescein dye examination were performed prior to randomisation to ensure the selected rabbits had normal ocular health. Rabbits were distributed into two groups for both plain and microfabricated membranes; control or sham treated group (G1) of six rabbits (three males+three females) and PLGA treated (G2) group of 14 rabbits (seven males+seven females). Fibrin glue and a bandage contact lens were applied to the left eye of G1 rabbits. In G2 rabbits, PLGA membrane was placed on the left eye using fibrin glue, followed by the application of bandage contact lens. Before performing complete tarsorraphy, one drop of 2.5% povidone-iodine was instilled in the eye receiving the PLGA membrane or sham treatment. Surgical procedures were performed under general anaesthesia using a combination of ketamine (30 mg/kg body weight) and xylazine (5 mg/kg body weight). After removal of the bandage contact lens, ciprofloxacin eye drops were administered four times a day for 1 week and carboxymethyl cellulose drops applied daily. Four rabbits (two male+two female) each, from G2, were sacrificed on day 8 and day 15 to study the degradation of the membranes in vivo. All remaining animals were sacrificed on day 29.

A schematic of the experimental protocol is given in figure 1.

Figure 1

Schematic of the experimental protocol in rabbits. This protocol was used for both the study of the plain poly-lactic co-glycolic acid (PLGA) membrane and of the PLGA membrane containing microfabricated pockets which were conducted as two separate experiments with individual control groups.

Test material

PLGA membranes, 50 µm thick with fibre diameters of 2–3 µm were electrospun.9 Polymers were dissolved in 1,1,1,3,3,3-hexafluoroisopropanol (Sigma Aldrich, Dorset, UK) at a concentration of 20 wt%. Polymer solutions were delivered at a constant feed rate of 800 µL/hour, using a programmable Harvard PHD4400 syringe pump (Harvard Apparatus, Kent, UK) to a blunt tipped stainless-steel needle with an internal diameter of 0.8 mm. The tip of the needle was in turn connected to a positive high voltage unit (Glassman High Voltage High Bridge, New Jersey, USA) and solutions were electrospun with an applied voltage of 12.5 kV. Electrospinning was performed in an environmentally controlled, A1-Safetech, air recirculation cabinet at 25°C and a relative humidity of 25%. For the plain PLGA membranes, fibres were deposited onto a grounded, custom built rotating mandrel at a distance of 300 mm from the tip of the needle, coated in aluminium foil. For the membranes with microfabricated pockets, fibres were deposited onto templates prepared via stereolithography.14

Following production and quality control assessment by scanning electron microscopy (Phenom G2 Pro, The Netherlands) to determine membrane thickness and fibre diameter, membranes were dried under vacuum at room temperature for 48 hours to remove residual solvent, before being die cut into 22 mm diameter discs. The membranes were then vacuum sealed in bags prior to terminal sterilisation via γ-irradiation at Steris AST (Moray Road, Swindon, UK). A schematic representation of the production process is given in figure 2.

Figure 2

Schematic of the production and sterilisation route for poly-lactic co-glycolic acid (PLGA) electrospun membranes (plain and microfabricated). The membrane design (top panel) involved the fabrication of the membranes in collaboration with The Electrospinning Company which ensured both reproducibility and scaling-up possibilities. For the production of the microfabricated membranes, first polymeric ring templates of 1.6 cm of diameter were created using a layer-by-layer photocuring approach in which a blue laser (473 nm) was directed into a bath of polyethylene glycol diacrylate containing camphorquinone during intervals of 15–60 s. The process involved the use of a computer-aided design comprising two layers (layer 1 acted as a base and it did not contain microfeatures; layer 2 contained a total of six horse-shaped microfeatures of sizes ranging 350–500 µm). The combination of these two layers resulted in the creation of the polymeric rings which acted as templates. These were mounted onto a 2 mm thick grounded electroplated aluminium sheet and then membranes were electrospun over them. This resulted in microfabricated PLGA electrospun rings (pocket membranes). The manufactured membranes were characterised and tested in vitro using both rabbit and human corneal cells. After product design optimisation and in vitro testing, prototype membranes were produced and sterilised using gamma irradiation. PLGA membranes were then vacuum packed with a range of desiccants to ensure moisture stability. The optimised and sterilised membranes (plain and microfabricated) were then used for the evaluation of local and systemic toxicity in rabbits.


Clinical signs

Clinical examinations were performed prior to treatment and weekly thereafter. These included changes to skin, fur, eyes, mucous membrane, secretions, excretions, autonomic activity, gait, posture and response to handling and the presence of clonic or tonic movements.

Body weight

Body weights were recorded on days 1, 8, 15, 22 and 29. Fasting body weights for interim sacrificed animals were taken 1 day prior to necropsy.

Ophthalmic investigations

  • Gross ocular observations were performed by visual assessment for signs of infection or inflammation. Treated eyes were examined for degradation of scaffold on days 8, 15 and 29 after removal of suture.

  • Fundus examination and IOP of all rabbits was carried out using a Kowa Digital Ophthalmoscope (Welch Allyn, USA) and a Reichert Tono-Pen XL Tonometer (Reichart, USA) prior to randomisation and sacrifice.

  • Fluorescein was applied to the ocular surface to assess health prior to randomisation and on sacrifice days after removal of suture material, contact lens and PLGA membrane where applicable.

Blood investigations

Clinical laboratory investigations were performed on days 8, 15 or 29. Blood was collected from the overnight fasted rabbits from the marginal ear vein. Haematological parameters were determined using ADVIA, 2120, Siemens, Germany.

Clinical chemistry parameters were analysed with the help of an automatic biochemical analyser (Vitros 250, Johnson and Johnson) using standard kits.

Tissue investigations

The animals were fasted overnight, weighed and euthanised under over anaesthesia of thiopentone sodium.

Organ weights

Adrenals, brain, heart, lungs, kidneys, liver, spleen, testes/uterus and thymus were dissected free of fat and weighed wet as soon as possible to avoid drying.


Tissues from sacrificed rabbits were processed in an automatic tissue processor (Microm spin tissue processor STP 120-3, Thermo Scientific, USA) and embedded in paraffin wax using Microm Embedding centre EC350, Thermo Scientific, USA. Tissue sections of 3–6 μm were stained with H&E in an automatic tissue stainer (Microm HMS 740 robotic stainer, Thermo Scientific, USA). A four-step grading system of minimum, mild, moderate and marked was used to rank microscopic findings for comparison among groups.

Statistical analysis

The data generated from the rabbits subjected to interim sacrifice on day 8 and 15 was not considered for statistical analysis along with the animals sacrificed on day 29 (due to variance in the day of sacrifice compared with control animals). The data is expressed as mean±SD Normality of the data was confirmed using D’Agostino and Pearson omnibus test. Analysis of variance (ANOVA) and Kruskal-Wallis tests were used for group comparison of non-homogenous data. Parameters showing significance in the Kruskal-Wallis test were further analysed with Wilcoxon’s test to compare each group individually over the respective control arm. The data analysis was performed using SAS V.9.2, Enterprise Guide V.4.2. All statistical tests were performed at 5% level of significance.


Rabbits from control and PLGA-treated groups did not show any treatment related abnormal clinical signs during the experimental period. There were no deaths in any of the groups of rabbits.

Body weight

No statistical differences were observed in the mean body weights (online supplementary table 1) of the different groups of rabbits throughout this study.

Ophthalmic examination

The PLGA membranes (plain and microfabricated) appeared intact on the ocular surface on day 7. Partial breakdown of the membrane was noted in all animals on day 15. By day 29, there was complete breakdown of the respective membranes in all animals (figure 3).

Figure 3

Breakdown of plain and pocket PLGA membranes on corneas of rabbits. Representative images of the residual plain PLGA membrane on the ocular surface of the rabbit on days 8,15 and 29 is shown in panel A. Representative image of the sham treated eye on day 29 is shown for comparison. In panel B are shown representative images of the ocular surface of control eyes and eyes treated with PLGA membranes after fluorescein staining on days 8, 15 and 29. The staining seen in these images are of the PLGA membranes that remained attached to the ocular surface. Panel C is a summary table of the ocular findings in the animals treated with plain or pocket PLGA and their respective control animals.

Mild to moderate conjunctival congestion was noted in all animals sacrificed on day 7 but resolved completely by day 29. In all animals sacrificed on day 7 (both plain and microfabricated), PLGA membrane was found adhered to the ocular surface. Vascularisation (<2 clock hours) was noted in 1/4 and 3/6 animals that received the plain PLGA membrane at the end of 15 and 29 days, respectively. Control animals showed epithelial defects and stromal thinning in 2/6 (day 15) and 1/6 (day 29) animals, respectively. Corneal vascularisation (2–4 clock hours) was noted in 3/6 animals of the pocket PLGA-treated animals and in 3/6 of control animals at the end of the 29 days.

There was no significant change to the IOP when compared with the control animals (online supplementary figure 1). This was true for both membranes with and without microfabricated pockets. Similarly, no abnormalities were detected in the retina of any of the animals before or after treatment with the PLGA membranes. A summary of the ophthalmic observations is given as a table in figure 3.


In animals that received the plain membrane, a significant decrease in mean corpuscular haemoglobin concentration (p≤0.01), RBC distribution width (p≤0.05), relative and absolute monocyte counts (p≤0.05) and absolute lymphocyte counts (p≤0.05) was observed compared with control group. A significant increase in relative basophil count (p≤0.05) was also observed in rabbits treated with the test membrane compared with the control group (table 1).

Table 1

Haematological values

In animals that received the pocket membranes, a significant increase in absolute lymphocyte counts (p≤0.05) was observed in rabbits treated with the test items compared with the control animals. No other parameter varied significantly between the two groups (table 1).

Clinical chemistry

Alkaline phosphatase (ALP) (p≤0.05) and gamma-glutamyl transferase (GGT) (p≤0.01) were significantly increased in rabbits treated with plain membranes compared with control rabbits and there was a significant decrease in creatinine (p≤0.05) and total bilirubin (p≤0.01) values when compared with control rabbits (table 2).

Table 2

Clinical chemistry values

In animals that received pocket membranes, there was a significant decrease in alanine aminotransferase (ALT) values (p≤0.05) and a significant increase in blood urea nitrogen (BUN) (p≤0.01) values when compared with control rabbits. All other parameters were comparable in both groups (table 2).

Organ weights

There were no significant changes in the absolute or relative organ weights in male and female rabbits that received the plain membrane compared with the control animals.

In the animals with the pocket membranes, absolute (p≤0.01) and relative (p≤0.05) weights of the liver were significantly higher in the treated group compared with the control group (online supplementary table 2).

Gross pathology

There were no gross pathological findings indicative of inflammation or immune response in any of the rabbits treated with plain or pocket membranes or with sham treatment.


Hyperplasia of corneal epithelium and remodelling of anterior stromal collagen fibres were noted in 1/6 control animals (sacrificed on day 29) of the plain membrane group (online supplementary table 3). The same was noted in 2/6 control animals and 4/6 animals that received the pocket membrane (day 29; online supplementary table 4). Three rabbits that had the pocket membrane showed infiltration of heterophils at the filtration angle (online supplementary table 4).

There were no lesions observed in the retina in any of the animals examined irrespective of whether animals received PLGA membranes or not.

In the liver, focal chronic inflammation was noted in 1/6 (plain PLGA) and mineralisation of hepatocytes in 1/6 (pocket PLGA) animals. Basophilic tubules and mineralisation were noted in the kidneys of 1/6 animals that received the plain and pocket PLGA membranes, respectively. Increased alveolar macrophages were noted in 1/6 animals that received both plain and pocket PLGA membranes. All other histological changes were comparable between the control and PLGA-treated animals.

Discussion (The discussion is not the same as in the revised submission of manuscript (Submitted in July 2018). In the revised manuscript, the citations were also more (21). Please check the version of the manuscript that the editor has approved for publication)

PLGA has been used for many years for a wide range of applications. It degrades to polylactic and polyglycolic acid without causing any systemic toxicity largely driven by hydrolytic mechanisms rather than enzymatic breakdown.8 ,10 ,15 ,16 17

It has been used in delivering anti-inflammatory agents following cataract surgery15; implanted in subconjunctival space of rabbits for 6 months18 and to support cells with minimal cell toxicity.19 We previously demonstrated that limbal stem cells could be cultured on these PLGA membranes where they provided support comparable to the hAM.9

Prior to a first in man safety study, it was necessary to show that these membranes would not induce local or systemic toxicity when applied to the cornea of a relevant animal model. The results we obtained showed that membranes degraded within 29 days without causing significant topical or systemic toxicity.

While temporary conjunctival congestion was noted in all animals on day 7, this was a response to surgical procedures. Vascularisation of the corneal surface (2–4 clock hours) was noted in both sham and PLGA-treated animals indicating the response was primarily to the surgical procedure. Critically, fluorescein staining showed clear and intact ocular surface in all animals except one control animal that had stromal thinning. Haematological parameters showed significant but minor changes in lymphocyte and monocyte counts in the test animals—both decreases and increases but all within the normal expected range of clinical values for these animals. As the gross and individual organ histology findings did not indicate sustained inflammation or infection, these changes were considered to be adaptive, consistent with the expected immune response to the breakdown of the PLGA material.

Similarly, while there were significant decreases in the creatinine and total bilirubin levels in plain PLGA-treated animals, significant increases in ALP and GGT values, an increase in BUN values and a decrease in the ALT values in the pocket PLGA-treated animals compared with the control animals, these were again within the normal range for these animals and there was no histological evidence of any adverse reaction to the PLGA membranes.

There was a significant increase in the liver weight for animals that received the pocket membrane but again no evidence of any correlation with clinical histopathological findings as would occur with hepatocellular hypertrophy which can occur in response to toxicity. We suggest the increase in the liver weight in all the animals of pocket PLGA-treated animals might be indicative of a mild inflammatory response. In an earlier study,10 in which we implanted electrospun sheets of various ratios of polylactic acid to polyglycolic acid into the flanks of rats, we found a vigorous macrophage response associated with scaffold breakdown but little evidence of any lymphocytic response.

In conclusion, the toxicity studies showed no evidence of topical or systemic toxicity in response to the placing of these membranes on the cornea of rabbits.


We acknowledge the Wellcome Trust (Affordable Healthcare in India) for the funding support. We are thankful to Dr Jomy Jose from Vimta Labs Ltd for her help with the interpretation of the animal data.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.

Supplementary materials

Related Data


  • Contributors CR was responsible for the conception of study, data analysis, interpretation of data and manuscript preparation. VSS was involved in the conception of study and interpretation of data. IO was responsible for the design of PLGA materials and manuscript preparation. UB and SM were involved in the design of study, data collection and analysis. RMK contributed to the manuscript preparation. SMN contributed to the conception and design of study, data interpretation and manuscript preparation and approval.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Ethics approval Institutional Animal Ethics committee of VIMTA Labs Limited.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost B Jonas