A Low-Molecular-Weight Oil Cleaner For Removal of Leftover Silicone Oil Intraocular Tamponade

Abstract

Purpose.: Silicone oil (SO) has been used as a long-term intraocular tamponade in treating retinal diseases for more than half a decade. However, its propensity to form tiny SO droplets is associated with a number of complications. Currently there is no effective way to remove such droplets from the eye cavity. In this work, a novel cleaner was developed for effective removal of these droplets.

Methods.: The cleaner promotes the formation of an oil-in-water-in-oil (O/W/O) double-emulsion that consists of the unwanted droplets as the innermost oil phase. The cleaner's ability to encapsulate SO droplets was tested using both in vitro microdevices and ex vivo porcine eye models. The efficiency of the cleaner in removing the SO droplets was quantified using the three-dimensional (3D) printed eye model. Both the volatility and in vitro cytotoxicity of the cleaner were evaluated on three retinal cell lines.

Results.: Cleaner 1.0 is volatile and has an evaporation rate of 0.14 mL/h at room temperature. The formation of O/W/O double-emulsion indicates the encapsulation of SO droplets by the cleaner. In the 3D printed eye model, rinsing with cleaner 1.0 led to a significant reduction of leftover SO droplets compared with 1× phosphate-buffered saline (PBS; P < 0.05; n = 6). Cleaner 1.0 did not cause significant cell death (3%–6%) compared with balance salt solution (BSS; 1%–3%) in all three cell lines. The reduction in the cell viability due to cleaner 1.0, relative to that of BSS, was significant only in ARPE-19 cells (27%; P < 0.05) but not in the other two cell lines (8% and 17%, respectively; P > 0.05).

Conclusions.: The double-emulsification approach was effective in removing remnant droplets from the eye cavities, and the cleaner was compatible with common cell types encountered in human eyes. The mechanism of toxicity of the proposed cleaner is still unknown, therefore, further in vivo animal tests are needed for full evaluation of the physiological response before the proposed cleaner can be advanced to clinical trials for retinal surgeries.

Introduction

Silicone oil (SO) with a high viscosity has been used as a long-term intraocular tamponade in vitreoretinal surgery for over 50 years.¹ It is commonly used in treating complicated retinal detachment,² giant retinal tear,³ ocular trauma,⁴ and proliferative vitreoretinopathy.⁵ These procedures often lead to formation of a stable interface between SO and the aqueous in the eye cavity and effectively restore the detached retina back to its underlying retinal pigment epithelium. This also helps to facilitate chorioretinal adhesion⁶ and prevent penetration of the aqueous through retinal holes to subretinal space.⁷ Due to its immiscibility with intraocular fluid and blood, SO remains transparent, and patients can see clearly through the SO after proper refraction adjustment.⁸ However, despite the in vivo biocompatibility of SO, the use of SO has been associated with various postoperative complications including glaucoma⁹ and inflammation, which can lead to proliferation, traction, and redetachment.¹⁰ These complications are highly correlated with the in vivo emulsification of SO, which results in numerous small SO droplets within the eye cavity.¹¹ In vivo emulsification of SO occurs due to the shear stress generated by repetitive eye movements at the interface between the SO phase and the aqueous phase.¹² There is no absolute agreement on the optimal time for the SO removal because the onset of SO emulsification depends highly on the individual clinical situation of the patient. To reduce the risk of SO-induced complications, recommended time ranges from 3 to 6 months^13,14; however, SO is often left in the eye for more than 12 months, unless secondary complications are found.¹⁵ Normally the time for SO removal is based mainly on the discretion of the vitreoretinal surgeons. 

Removal of SO from the eye cavity is typically performed by introducing two surgical openings 3.5 mm away from the limbus region.¹⁶ Balanced salt solution (BSS) is infused into the vitreous cavity through one of the openings while SO is aspirated through the other opening. Attempts have also been made to remove these SO droplets by repetitive flushing using BSS and air in a procedure called triple-fluid/air exchange, right after the removal of SO.² However, flushing of the eye with BSS, followed by air, is effective only in the early postoperative period but not later.¹⁷ Emulsified SO droplets are still seen in the anterior chambers of patients' eyes in the postoperative examination after the removal of SO. The complications due to SO emulsification cannot be prevented.¹⁸ One explanation is that the emulsified SO droplets are very stable inside the eye cavity, probably due to the stabilization by biomolecules, such as proteins and glycolipids inside the eye. These droplets often adhere on the surface of retina. The repetitive washing with BSS only dilutes the content of SO droplets inside the eye cavity but cannot achieve a complete removal. Therefore, to prevent postoperative complications due to the treatment, a more effective method for removing the SO droplets is greatly demanded. 

The goal of the current research is to develop a novel cleaner to reduce the intraocular complications induced by the emulsified SO droplets. Physically, an ideal cleaner should have a low viscosity, which offers the ease in both injection and aspiration. It should also be volatile in nature, ensuring that any leftover cleaner can leave the eye compartment through evaporation and diffusion into the blood capillaries. Biologically, it should be compatible with the retinal tissue and cause no adverse effect on the intraocular tissue during the washing procedure. Functionally, the cleaner should allow more effective removal of the emulsified SO droplets. 

In this study, we developed a novel cleaner to achieve a significantly higher efficiency in the removal of emulsified SO droplets than BSS in an ex vivo porcine eye model by encapsulating the emulsified droplets in the form of double-emulsion (Fig. 1).¹⁹ The proposed cleaner consists of a low-molecular-weight (LMW) SO (LMW-SO) with a hydrophobic surfactant dissolved in it. Its effectiveness in removing the emulsified droplets was demonstrated in both in vitro and ex vivo eye models. By introducing our cleaner, a very stable oil-in-water-oil (O/W/O) double-emulsion is formed, causing the water content adherent to the retinal surface to emulsify the embedded emulsified droplets. The unwanted SO droplets are therefore trapped inside the double-emulsion and can be easily removed. The volatility of the cleaner is high, allowing removal of any leftover oil by evaporation. The cleaner also exhibits an acceptable in vitro compatibility with the cells, similar in extent to various agents currently used in vitreoretinal surgery, which supports its use as an intraoperative agent in corresponding surgical treatment. 

Figure 1

 

Schematic flowchart of the use of the cleaner after the removal of SO from the eye cavity.

Materials and Methods

Formulations of the Cleaner

The proposed cleaner is a mixture of 95% LMW-SO and 5% silicone-based hydrophobic surfactant (v/v). The two cleaners, which we call cleaner 0.65 and cleaner 1.0, were made by mixing LMW-SO with hexamethyldisiloxane (Aldrich 469300, 0.65 mPa·s; Sigma-Aldrich Corp., St. Louis, MO, USA) and octamethyltrisiloxane (Aldrich 469319, 1.0 mPa·s; Sigma-Aldrich Corp.) with a hydrophobic surfactant (Dow Corning fluid 749 [DC749]; Dow Corning Corp., Midland, MI, USA) respectively. DC749 fluid is a commercial grade product. The physical properties and components of the two types of LMW-SO and the surfactant DC749 are listed in the Table. 

Table

 

Physical Properties of the Components in the Cleaner

Agent	Silicone oil 0.65 mPa·s	Silicone oil 1.0 mPa·s	DC749 Fluid
Nature	LMW-SO	LMW-SO	Hydrophobic surfactant
Composition	Hexamethyldisiloxane	Octamethyltrisiloxane	50% Decamethylcyclopentasiloxane & 50% Tetra(trimethylsiloxy)silane
Viscosity (25°C), mPa·s	0.65	1.0	500
Density (20°C), g/mL	0.764	0.82	1.05
Melting point, °C	−59	−82	Unknown
Boiling point, °C	101	153	210
Refractive index	1.377	1.384	1.405

O/W Emulsion Generation

An O/W emulsion was prepared using high-shear homogenization (IKA T10 Basic ultra-Turrax, IKA-Werke GmbH & Co. KG, Germany), with 99% phosphate-buffered saline (PBS) and 1% silicone oil, 1300 mPa·s (v/v). The O/W emulsion prepared by homogenization was used as a model postoperative intraocular fluid. 

Studies With an In Vitro Eye-Like Microdevice

The device was fabricated to mimic the sagittal plane of a human eye. The device consisted of three layers, with the geometry of the eye cavity in the middle layer. The eye geometry was developed using laser engraving. The inner wall of the layer simulating the geometry of the eye was coated with 5% bovine serum albumin (USB Corp., Cleveland, OH, USA) to mimic the surface property of the human retina.²⁰ The device was first filled with the O/W emulsion prepared by homogenization for 2 hours to allow emulsified droplets to adhere to the inner wall of the device. Residual emulsion inside was subsequently removed. After that, the cleaner candidate was infused to encapsulate the remaining aqueous together with unwanted SO droplets. 

Ex Vivo Porcine Eye Studies

Porcine eyes were freshly harvested from a local slaughterhouse. Three port pars plana vitrectomy procedures were performed on the porcine eyes, using a vitrectomy machine (Accurus; Alcon, Fort Worth, TX, USA). Surgical instruments including a vitrectomy probe and an endoilluminator were introduced into the eye cavity though the pars plana, and the vitreous humor was removed using vitreous cutter. After vitreous was removed, the eyes were filled with an O/W emulsion. The O/W emulsion generated by homogenization was kept inside the porcine eyes for 2 hours to allow emulsified droplets to adhere to the tissues of the eye cavity. O/W emulsion was removed by fluid-air exchange afterward. The cleaner was infused to encapsulate the remaining aqueous together with unwanted SO droplets. Images of the vitreous cavity were captured from the anterior side of the eyeball through the cornea by using a surgical microscope. The experiment was repeated twice. 

Three-Dimensional (3D) Printed Eye Model Experiment

The 3D hollow eye models were fabricated using a rapid prototyping machine (Objet Eden350V; Stratasys Ltd., Eden Prairie, MN, USA). The inner walls of the 3D eye models were coated with 5% bovine serum albumin to mimic the surface property of the human retina. This O/W emulsion prepared by homogenization was injected into the 3D eye models and kept inside for 2 hours to allow emulsified droplets to adhere to the inner surface of the 3D eye models. The cavities of the models were then washed by the proposed cleaner. The washed models were then placed inside an oven at 37°C for 24 hours to allow any remaining cleaner to be evaporated. Afterward, PBS was used to wash the cavity, and the residue was collected and analyzed using a particle counter (Multisizer 4; Beckman Coulter, Inc., Brea, CA, USA) to quantify the number of droplets remaining in the cavity of the 3D models after the rinsing procedure. The rinsing efficiency was compared with that of the control experiment in which only PBS was used as the washing agent. The sample size of each group equaled six (n = 6) for this experiment. 

Evaporation Rate Studies

The evaporation rates of the LMW-SOs were measured by monitoring the change in mass of the samples, using an electron beam balance (CP225D; Sartorius AG, Göttingen, Germany) over time. One milliliter of the reagent was added to a 60-mm-diameter culture dish, and its initial mass was recorded. Mass of the testing samples was plotted against time. Evaporation rates of the two cleaners, the two pure LMW-SOs, and the pure surfactant DC749 were measured. 

In Vitro Biocompatibility Tests

The cytotoxicity effect of the cleaner candidates on the cells was tested using an in vitro transwell culturing system.²¹ Three retinal cell lines, including a retinal ganglion cell line (RGC-5), a retinal Müller cell line (rMC-1), and a retinal pigment epithelium cell line (ARPE-19), were used. RGC-5 was a type of rat retinal ganglion cell (RGC-5; ATCC, VA, USA) originally derived by transforming postnatal rat retinal ganglion cells by using virus.²² RGC-5 cells at passages 11 to 13 were used. rMC-1 is an immortalized rat Müller cell (rMC-1) originating from adult rat retina tissue.²³ The rMC-1 at passages 30 to 32 were used. ARPE-19 cells were immortalized from human retinal pigment epithelial cells.²⁴ RGC-5 at passages 11 to 13 were used. Both the RGC-5 and the rMC-1 cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco, Life Technologies Corp., Carlsbad, CA, USA) with 10% fetal bovine serum (Gibco), 100 U/mL penicillin and 100 μg/mL streptomycin (Gibco) as supplements. ARPE-19 cells were maintained in DMEM-F-12 medium (Gibco) with 10% fetal bovine serum (Gibco), 100 U/mL penicillin, and 100 μg/mL streptomycin (Gibco) as supplements. The cells were cultured as monolayers on 12 mm-diameter semipermeable, 0.4-μm-pore-size polyester filters (Costar Transwell; Corning, Inc., Corning, NY, USA). Purified mouse laminin (Sigma-Aldrich) diluted in BSS was used to coat the apical side of the filter with a concentration of 2 μg/cm² at 37°C for 6 hours. Each filter was seeded with 2.5 × 10⁴ cells on the apical side; and 0.5 mL and 1.5 mL of the medium was added to the apical compartment and the basal compartment, respectively. After cells were grown on the filters for 24 hours, the medium was removed from the apical side. There were eight testing groups BSS, perfluorodecalin (PFD),²⁵ perfluorohexyloctane (F6H8),²⁶ hydrophobic surfactant (DC749), 0.65 mPa·s LMW-SO, rinse 0.65, 1.0 mPa·s LMW-SO, and rinse 1.0. Groups tested in BSS were treated as the controls in this work. PFD and F6H8 are two widely accepted short-duration intraoperative adjunct compounds used in retinal surgeries and were used as the control samples with acceptable cell viability. Cleaner candidates based on two LMW-SOs (0.65 mPa·s and 1.0 mPa·s) with 5 vol% of DC749 were tested. Cells were incubated with 0.3 mL of the various testing agents for 1 hour. The filter was covered with a glass cover slip to prevent evaporation of the testing agents during the incubation period. After the 1-h incubation, the testing agents were removed. BSS was added to the apical side of the well and kept for 24 hours. CellTiter96 AQueous nonradioactive cell proliferation (MTS) assay (Promega, Dane County, WI, USA) and cytotoxicity detection kit (LDH) assay (Roche, Basel, Switzerland) were then used to study relative cell viability and cell death, respectively, by using the colorimetric method. The resultant readings of both the MTS and LDH assays were means from three samples. Statistical significance was assessed using the statistical test of one-way analysis of variance (ANOVA), followed by a post hoc Bonferroni test. A P value of <0.05 was considered statistically significant. The morphologies of the cells under different testing agents before and after the experiments were captured under light microscopy. All the experiments were repeated three times for each cell line. 

Results

Encapsulation of Emulsified Silicone Oil Droplets by Using the Cleaner

After cleaner was infused into the in vitro microdevice, the O/W emulsion, which was used as the model for leftover SO droplets in the eye, was encapsulated inside some O/W/O double-emulsion droplets (Fig. 2A). A similar result was performed in the ex vivo porcine eye (Fig. 2B). The newly formed doubly emulsified aqueous globules were found to form a sediment in the lower part of the microdevice (Fig. 2 [A2]). 

Figure 2

 

(A) Formation of O/W/O double emulsion within the in vitro eye-like microdevice after infusing the proposed cleaner into the aqueous-containing cavity. (A1) Design of the microdevice that mimics the geometry of the sagittal plane of the eye cavity. (A2) Encapsulation of SO droplets (red arrows). Black bars: 200 μm. (B) Formation of O/W/O double emulsion within the vitreous chamber of the ex vivo porcine eye after infusing the proposed cleaner into the vitreous cavity. (B1) Setup of the vitrectomy. (B2) Double emulsion in the vitreous cavity. The red arrows indicate the double-emulsified droplets. Red bars: 200 μm.

Effectiveness of the Cleaner in Removing Emulsified SO Droplets

The effectiveness of the proposed cleaner was quantified by counting the number of droplets in the washout, using the 3D-printed hollow eyeball model (Fig. 3A). The number of droplets remaining in the 3D hollow eyeballs after flushing with the cleaner was significantly lower than that with PBS in the controls (300 vs. 950 droplets, respectively; P < 0.05, n = 6), as indicated by the droplet count using a Coulter counter (Fig. 3B). 

Figure 3

 

(A) 3D printed hollow eyeball models. The diameter of the 3D printed eyeballs is approximately 2.5 cm. There are two openings located at the limbus region (arrows). (B) The percentage of oil droplets remaining in the 3D printed hollow eyeball after the washing procedure. There are significantly fewer SO droplets remaining in the cleaner-washed eyeball models than in PBS-washed models, demonstrating the efficiency of the cleaner in this application. Unpaired t-test; *significant differences from control PBS; P < 0.05; error bar: + SD; n = 6.

Evaporation Rate of Cleaner Candidates

To confirm their ease of removal, we measured the evaporation rates of LMW-SO 0.65 mPa·s and 1.0 mPa·s, which were 1.33 mL/h and 0.14 mL/h, respectively (Fig. 4). DC749 has a very low evaporation rate, with less than 0.01 mL/h. Cleaner 0.65 and cleaner 1.0 had rates of evaporation similar to their base LMW-SOs, 0.65 mPa·s and 1.0 mPa·s, respectively. 

Figure 4

 

Evaporation rates of LMW-SO, cleaner candidates, and surfactant DC749. Cleaner 0.65 has an evaporation rate 9 times faster than that of cleaner 1.0.

Cytotoxicity of Cleaner Candidates

Cleaner 1.0 had no significant adverse effects on the retinal cell lines. There was no statistically significant difference in the cell viability (Fig. 5A) and cell death (Fig. 5B) of rMC-1 and RGC-5 cells between the cleaner 1.0, the control BSS or PFD (P > 0.05). The result was also consistent with the morphological study (Figs. 6A, 6B). There was a 27% decrease in cell viability of ARPE-19 cells after incubating with the cleaner 1.0 relative to BSS group (P < 0.05). However, this decrease in cell viability was not as significant as that due to F6H8 (66%) (P < 0.05) (Fig. 5A). In addition, we could not see any significant difference in the cell morphology of the ARPE-19 cells between the cleaner 1.0 and the control samples of BSS and PFD (Fig. 6C). Also, there were no significant adverse effects on the cells by incubating with pure LMW-SO 1.0 mPa·s and pure DC749 compared with BSS (P > 0.05), which are the components of cleaner 1.0. 

Figure 5

 

Results of the cell viability (A) and cell death (B) tests 24 hours after the 1-hour incubation with various testing agents. One-way ANOVA test followed by Bonferroni multiple comparison test; *significant differences from control BSS; P < 0.05; error bar: ±SD; n = 3.

Figure 6

 

Cell morphologies of rMC-1 (A), RGC-5 (B), and ARPE-19 (C) cells 24 hours after the 1-h incubation with various testing agents. Significant cell death is observed in both the 0.65 and 0.65 mPa·s cleaner groups. F6H8 also induced cell death in ARPE-19 cells. Red bars: 100 μm.

On the other hand, cleaner 0.65 showed poor compatibility with the cells. Cells incubated with cleaner 0.65 had significant decrease of cell viability (70%; P < 0.05) (Fig. 5A) and significant cell death (15%–25%; P < 0.05) (Fig. 5B) compared with BSS control, PFD, and cleaner 1.0. A large number of floating cells with circular shapes were observed, indicating the cytotoxic effect cleaner 0.65 had on the cells (Figs. 6A, 6B, 6C). Similar results were obtained in the LMW-SO 0.65 mPa·s group (P < 0.05) but not in the DC749 group (P > 0.05) when compared with BSS. 

Discussion

The cleaner is capable of encapsulating emulsified SO droplets both in vitro and ex vivo (Figs. 2A, 2B). According to the Bancroft rule, “the phase in which a surfactant is more soluble constitutes the continuous phase.”^27,28 The properties of relative hydrophobicity and amphilicity of the surfactant used (DC749) facilitate the formation of W/O emulsion. By introducing DC749 into the LMW-SO, the water adherent to the retinal surface was emulsified. The double-emulsion does not require high shear to form; based on our experimental data, simple flushing of a model eye cavity with the proposed cleaner led to successful formation of double emulsions. Moreover, these double emulsions formed sediments in the lower part of the microdevice (Fig. 2A [A2]) due to the density differences between cleaner and aqueous phases. LMW-SO 0.65 mPa·s and 1.0 mPa·s have the specific gravity of 0.76 g · cm⁻³ and 0.82 g · cm⁻³ respectively. The cleaner has a lower density than water; thus the aqueous globules with emulsified SO droplets encapsulated sink. These globules can then be easily collected and removed. 

We quantified the effectiveness of the proposed cleaner in removing the unwanted SO droplets by counting the number of droplets in the washout. Ex vivo porcine eye is currently the best physical model to mimic the intraocular environment. However, during the delicate and time-consuming procedures of vitrectomy, tissue and cell debris inevitably appear. Moreover, intraocular tissues deteriorate after harvesting. Both effects result in increased particle counts in the washout, making the droplet counting infeasible. To tackle this problem, we use a 3D printed hollow eyeball model (Fig. 3A) as the model for this experiment. Droplets found in the washout does not contain any tissue or cell debris, and can be attributed to the emulsification of SO added only. We show that our proposed rinse has a higher ability in removing the emulsified SO droplets than the control PBS (Fig. 3B). This result further supports the significantly enhanced efficiency of cleaner in droplet removal. 

During surgical procedures, it is often difficult to ensure complete removal of the cleaner. It is therefore important to consider a way for the remaining cleaner to leave the eye cavity; otherwise, this runny liquid would be dispersed easily and form new and additional emulsified SO droplets inside the eye. These droplets might then migrate to the subretinal layers and cause damage to the retina or even to the optic nerve. LMW-SO, the major component of the cleaner, has been proposed to leave the eye cavity through diffusion from ocular tissue.^29,30 The small dose of the silicone oil with a low molecular weight did not show any harm in rats and guinea pigs.³¹ However, not all LMW-SO can leave the eye through evaporation. LMW-SO with viscosity higher than 1.0 mPa·s is not suitable for this application because of the impractically long evaporation time. From our measurements, LMW-SO 5.0 mPa·s does not evaporate for at least 8 hours. Therefore, in this study, we chose two of the smallest commercially available LMW-SOs. The evaporation rates of LMW-SO 0.65 mPa·s and 1.0 mPa·s, which are 1.33 mL/h and 0.14 mL/h, respectively (Fig. 4), are fast, and therefore any remaining cleaner should be capable of leaving the eye through diffusion into the capillary system and subsequent evaporation. The hydrophobic surfactant DC749 has a very low evaporation rate (less than 0.01 mL/h), but any remaining DC749 can be washed away simply by constant injection of LMW-SO due to its high solubility in LMW-SO. It is also feasible to fill the eye cavity with a long-acting gas such as SF₆ and C₃F₈,³² after removal of the cleaner to facilitate evaporation of the remaining cleaner. 

The proposed cleaner should not have any adverse effect on retinal tissue in vivo. Because the washing procedure, which involves the use of the cleaner, should take a short period of time (e.g., less than 15 minutes) in the surgical procedure, we therefore studied the cell cytotoxicity effect of the cleaner candidates after short contact with the cells (1 hour). From our data, cleaner 1.0 had no significant adverse effect on the retinal cell lines. There also was no significant adverse effect on the cells by incubation with pure LMW-SO 1.0 mPa·s and pure DC749, which are the components of cleaner 1.0. These results suggest that cleaner 1.0 does not induce severe cytotoxicity to the cells in vitro, at least in the short term. Although the hydrophobic surfactant DC749 has been shown to be compatible with the relevant cell types in vitro, our approach was not limited to DC749. In theory, any biocompatible hydrophobic surfactant capable of inducing formation of W/O emulsion droplets should be applicable. On the other hand, cleaner 0.65 showed poor compatibility with the cells. This implies the toxicity of LMW-SO 0.65 mPa·s in vitro even in procedures with a short duration. 

LMW-SO 0.65 mPa·s and LMW-SO 1.0 mPa·s have the smallest molecular weights among all SOs tested. Although they are similar in physical properties, they exhibit different in vivo behaviors. LMW-SO 0.65 mPa·s causes a significantly more severe and rapid ocular edema and opacification than LMW-SO 1.0 mPa·s on the rabbit cornea over a period of 1 week.³⁰ The lipophilic LMW-SO 0.65 mPa·s molecules are expected to have entered the retinal tissues by dissolving in the lipophilic cell membranes,³⁰ thereby causing adverse effects.²⁹ Because the diffusion coefficients of the SO decrease with increasing molecular weight,³³ LMW-SO 0.65 mPa·s, which has a molar mass of 162.38 g/mol, can diffuse into the cells at a faster rate than LMW-SO 1.0 mPa·s with a molar mass of 236.53 g/mol, thus leading to a more severe adverse effect on the tissues. The significantly higher compatibility of the LMW-SO 1.0 mPa·s also suggests the existence of a critical molecular size, which is comparable to the molecular size of the LMW-SO 0.65 mPa·s, for entry into the cells. In this study, we showed that LMW-SO 1.0 mPa·s is compatible with the relevant cells in the retina, at least after contact for a short duration (less than an hour) in vitro. However, the compatibility of the rinse candidate with corneal endothelial cells has not been tested. Moreover, the in vivo biocompatibility of the cleaner remains to be investigated before clinical trials can be carried out. 

Conclusions

In this study, we introduced the novel cleaner 1.0, which is a mixture of LMW-SO and a hydrophobic surfactant, for removing unwanted SO emulsified droplets within the eye cavity by using a double emulsification approach. The novel cleaner 1.0 has three important physical properties that make it suitable for the desired application: the low viscosity facilitates the ease of both injection and removal; the low density of the cleaner ensures the doubly emulsified aqueous globules sediment to the lower part of the eye cavity where they can be easily collected and removed; and the high volatility allows any remaining cleaner to vaporize and leave the vitreous cavity. In addition, cleaner 1.0 does not cause severe cytotoxicity to the retinal cells in vitro and is comparable to or even better than two currently approved intraoperative tools tested, namely PFD and F6H8. Despite its promise, this approach is qualitatively new for removal of SO droplets. It is still in the research phase and is not yet ready for use in human studies. The long-term in vivo toxicity of the cleaner and its clinical effectiveness needs to be confirmed. Further investigations on in vivo biocompatibility of the cleaner are therefore needed before it can proceed to clinical trials to be used as a new intraoperative tool. 

Acknowledgments

The authors thank Charles Whitford, PhD, University of Liverpool, for providing the stereolithography drawing of the 3D eyeball model. 

Presented in part at the annual meeting of Association for Research in Vision and Ophthalmology, Seattle, Washington, United States, May 5–9, 2013. 

Supported by the Innovation and Technology Fund (ITS/105/13, InP/319/13, InP/31/13). Partially supported by the Early Career Scheme (HKU 707712P) and the General Research Fund (HKU 719813E and 17304514) from the Research Grants Council of Hong Kong, as well as the General Program (21476189/B060201) and Young Scholar's Program (NSFC51206138/E0605) from the National Natural Science Foundation of China. 

Disclosure: Y.K. Chan, P; D. Wong, P; H.K. Yeung, None; P.K. Man, None; H.C. Shum, P