Silicone oil adhesion to intraocular lenses: An experimental study comparing various biomaterials

Silicone oil adhesion to intraocular lenses: An experimental study comparing various biomaterials David J. Apple, MD, Robert T. Isaacs, MD, David G. Kent, FRACO, Louis M. Martinez, PhD, Soohyung Kim, MD, Stephanie G. Thomas, MD, Surendra Basti, MD, Derek Barker, MS, Qun Peng, MD ABSTRACT Purpose: To perform an in vitro experimental study comparing the degree of adherence of silicone oil to various rigid and foldable intraocular lens (IOL) designs and to the human lens capsule. Setting: Center for Research on Ocular Therapeutics and Biodevices, Department of Ophthalmology, Storm Eye Institute, Medical University of South Carolina, Charleston, South Carolina, USA. Methods: Seven IOL styles comprising various biomaterials were studied: fluorinetreated (FluorlensTM), heparin-surface-modified (HSM™), hydrogel, MemoryLens™, poly(methyl methacrylate) (PMMA), soft acrylic, and silicone lenses; the human crystalline lens was also studied. Each lens was immersed in silicone oil for 12 hours, then photographed, studied by scanning electron microscopy (except the crystalline lens), and subjected to computer-generated image analysis to determine the silicone oil coverage. Results: Silicone oil coverage of dry silicone lenses was 100% and of lenses immersed in normal saline, 82.5%. The least coverage was on the heparin-surface-

modified lens (mean score 9.4%). Coverage of the other four lenses ranged from approximately 15.1% to 33.7%. Mean coverage of the human lens capsule was 10.9%. Conclusion: Although a silicone IOL shows maximal adherence to silicone oil, other lens biomaterials are not immune to this complication. Silicone oil coverage was related to the dispersive energy component of the surface charge of the IOL biomaterial. Low dispersive energy materials had less silicone oil coverage, while those with higher dispersive energy had more oil coverage. J Cataract Refract Surg

1997; 23:536-544

A

dherence of silicone oil to an intraocular lens (lOL) may be harmful in cases in which silicone oil is used in retinal surgery in a pseudophakic patient. I - 4 In a previous study,1 we reported the clinically significant

Reprint requests to DavidJ Apple, MD, Department ofOphthalmology, Medical University ofSouth Carolina, 171 Ashley Avenue, Charleston, South Carolina 29425-2236, USA. 536

adherence of silicone oil used in vitreoretinal surgery to previously implanted silicone IOLs. Based on a study of three clinical cases, we determined that silicone oil should not be used in eyes in which a silicone IOL is already in place. The resulting opacity can obstruct the vitreoretinal surgeon's view into the eye intraoperatively and can also lead to significant visual loss for the patient. We cautioned against use of silicone IOLs in the rare

J CATARACT REFRACT SURG-VOL 23, MAY 1997

SILICONE OIL ADHESION TO 1015

patient with severe pre-existing vitreoretinal disease or with a high potential for vitreoretinal disease that would likely require intervention with silicone oil. It is of clinical significance to determine whether other 10L biomaterials may be susceptible to this complication. To accomplish this, we designed an experimental study in which we compared the degree of coating and adherence of silicone oil to rigid and foldable 10L designs made from silicone and other biomaterials. The differences revealed in this study may help a surgeon choose an appropriate 10L biomaterial for a patient with present or potentially severe vitreoretinal disease.

Samples of seven 10Ls were examined: (1) fluorine treated (Fluorlens™) (Chiron), (2) heparin surface modified (HSM™) (Pharmacia-Upjohn), (3) hydrogel (Storz), (4) MemoryLens™ (Mentor-Optical Radiation Corporation), (5) standard poly(methyl methacry-

late) (PMMA) (Pharmacia-Upjohn), (6) soft acrylic (Alcon and Allergan Medical Optics), and (7) silicone (Allergan Medical Optics, Staar Surgical, and Chiron Vision). Six 10Ls of each type were placed in silicone oil (AdatoSil5000™, Escalon Ophthalmics) for 12 hours overnight. After removal from the oil, each lens was photographed in air (dry state). Each lens was then immediately placed in water and gross photographs were obtained. Three silicone lenses were placed in normal saline for 12 hours before being placed in silicone oil. This was done to simulate the in vivo environment of the anterior chamber to see whether the silicone oil coating would be different on silicone lenses that had first been in an aqueous environment. Crystalline lenses from 12 human eyes obtained postmortem were placed for 6 hours in balanced salt solution and then for 12 hours in silicone oil that had been dyed red. The red dye was necessary because it was difficult to visualize the silicone oil on postmortem crystalline lenses, as opposed to the homogeneous, clear op-

A

B

Materials and Methods

Figure 1. (Apple) Silicone three-piece 10L (Allergan Medical Optics) coated with silicone oil. A: Gross photograph, showing almost 100% coverage. B: Image analysis, mean percentage coating (yellow area): 100%. C: Scanning electron micrograph, dry state, showing thick coating of silicone oil on both the optic and haptic (original magnification x8).

c J CATARACT REFRACT SURG-VOL 23, MAY 1997

537

SILICONE OIL ADHESION TO IOLS

B

A

Figure 2. (Apple) Soft acrylic IOL (Alcon) coated with silicone oil. A: Gross photograph showing two large foci and several tiny foci of coverage. B: Image analysis, mean percentage coating (yellow area): 33.7%.

tics of the IOLs. Red silicone oil was prepared by dissolving 0.1 g of Oil Red 0 (Fisher Biotech) in 5.0 cc of absolute ethanol. This mixture was stirred for 1 hour and then gravity filtered. Two milliliters of the filtrate was added to 5.0 mL of silicone oil and the mixture stirred for 6 hours at room temperature. The alcohol was then evaporated off at 60 to 70°e. Semiquantitative measurements of silicone oil coverage and qualitative assessments of adherence and mobility of oil on each IOL optic and crystalline lens were made. The semiquantitative comparative determinations were made using still photography and dynamic video analysis. Video analyses were performed with magnification provided by a Leica-Wild stereomi-

croscope (model M8). A computer-generated image analysis was performed on each IOL using Sigma Scan™ Measurement Software (Jandel Corp.) and SnappyTM Video System (Play Inc.) to determine the percentage of surface area covered by adherent silicone oil on the optic. After video analysis, the IOLs were examined by scanning electron microscopy (SEM). This analysis consisted of spuner-coating each specimen with gold and examining it at magnifications ranging from 8 X to 140 X, using the Cambridge scanning electron microscope (model number 360). To compare silicone oil with liquid perfluorocarbon (Fluorinert™, electronic liquid FC-43, 3M Industrial

A

B

Figure 3. (Apple) MemoryLens coated with silicone oil. A: Gross photograph. B: Image analysis, mean percentage coating (yellow area): 31.8%.

538

J CATARACT REFRACT SURG-VOL 23, MAY 1997

SILICONE OIL ADHESION TO 1015

A

B

Figure 4. (Apple) Standard one-piece all-PMMA IOL coated with silicone oil. A: Gross photograph. B: Image analysis, mean percentage coating (yellow area): 20.7%.

Chemical Products Division), another liquid material used as a tool in vitreoretinal surgery, we analyzed the reaction of each IOL immersed in the latter liquid.

Results Gross photographic analysis of one silicone lens style (A11ergan Medical Optics) showed maximum oil adherence (Figure I,A). The silicone oil could not be dislodged or moved by pressure from injected viscoelastic. Image analysis revealed 100% coating of the IOL optic in all cases (minimum 100%, maximum 100%, standard deviation [SD] 0%) (Figure 1, B). Scanning electron microscopy revealed a diffuse, thick coating over the optic and haptic of each IOL (Figure 1,C).

Identical findings were noted with Staar Surgical and Chiron Vision silicone IOLs. Silicone lenses that had first been placed in normal saline showed a mean silicone oil coverage of 82.5% (minimum 71.1 %, maximum 100%, SD 15.4%). Figures 2,A and B show gross photographic and image profiles of the three-piece soft acrylic IOL (Alcon AcrySof) after incubation in silicone oil. This profile was very similar to that seen with the MemoryLens (Figures 3, A and B). The Alcon and Allergan Medical Optics soft acrylic IOLs and the MemoryLens showed less tenacious oil adherence; mean percentage coating was 33.7% (minimum 15.0%, maximum 57.6%, SD 13.9%) and 31.8% (minimum 16.5%, maximum 50.8%, SD 12.2%), respectively (Figures 2,B and 3,B).

A

B

Figure 5. (Apple) HydrogellOL coated with silicone oil. A: Gross photograph. B: Image analysis, mean percentage coating (yellow area): 19.9%.

J CATARACT REFRACT SURG-VOL 23, MAY 1997

539

SILICONE OIL ADHESION TO IOLS

B

A

Figure 6. (Apple) Fluorine-treated IOl coated with silicone oil. A: Gross photograph. B : Image analysis, mean percentage coating (yellow area): 15.1%.

The standard PMMA IOL (Figures 4,A and B) and the hydrogel IOL (Figures 5,A and B) had silicone oil adherence that was similar to that on the soft acrylic and MemoryLens but somewhat lower image analysis values of 20.7% (minimum 5.3%, maximum 41.6%, SO 11.9%) and 19.9% (minimum 11.9%, maximum 40.2%, SO 10.5%), respectively. The fluorine-treated

lens (Figures 6, A and B) had a mean silicone oil coverage of 15.1% (minimum 9.1%, maximum 19.4%, SD 4.4%). Oil was readily removed with a viscoelastic. The heparin-surface-modified, all-PMMA IOL had the least silicone oil adhesion (Figure 7,A). Mean coverage was 9.4% (minimum 1.1 %, maximum 16.4%, SO 6.5%) (Figure 7, B). The oil droplets on the surface

B

A

Figure 7. (Apple) Heparin-surface-modified IOl coated with silicone oil. A: Gross photograph. B: Image analysis, mean percentage coating (yellow area): 9.4%. C: Scanning electron micrograph, dry state, showing a very thin, membrane-like coating on the optic and haptic (compare with Figure 1, C) (original magnification X15).

c 540

J CATARACT REFRACT SURG-VOL 23. MAY 1997

SILICONE OIL ADHESION TO IOLS

could be moved mechanically and dislodged from the optic surface. Scanning electron microscopy revealed some localized areas with a thin oil coating (Figure 7,C). There was little adhesion to 11 of the 12 human crystalline lenses. Oil coverage on 1 lens was 64.7%, almost three standard deviations greater than the mean. This extremely high value was probably caused by a problem with the lens or its processing and was therefore excluded from the study. The average coverage on the other 11 lenses was 10.9% (minimum 4.8%, maximum 27.2%, SD 6.4%) (Figure 8), and the oil was easily removed with a viscoelastic. With all seven IOL biomaterials, there was minimal liquid perfluorocarbon adherence. The adherence on the silicone IOL was 2.84% of the optic surface area

human lenses because it was impossible to visualize the transparent perfluorocarbon on the semiopaque crystalline lens. The mean percentages of silicone oil coverage on the seven biomaterials are shown in Figure 10. The percentage of oil coverage within each group ofbiomaterials and the lens capsules were found to be normally distributed. Pooled data were then analyzed in a multiple comparison method using a one-way analysis of variance (ANOYA) test. There was an extremely significant difference in silicone oil coverage between the groups of lens biomaterials (P < .001). Because coverage of the six silicone lenses was 100% in each case, with no variation between lenses, this group was used as a control to which the other groups were compared. The data were analyzed using the Bonferroni t-test, which is the most conservative test for multiple comparisons versus a control. There was no statistically significant difference in the extent of silicone oil coverage on silicone lenses that had first been placed in normal saline and those placed directly into silicone oil. There was, however, significantly less silicone oil coverage on each of the other lens biomaterials than on the silicone lens control group (P < .001). The power of the performed test was 1.000, indicating extremely high validiry of the results. The silicone oil coverage on the non-silicone IOL biomaterials and the lens capsule was then analyzed in a multiple comparison using a one-way ANOYA to determine whether significant differences existed. Data were again analyzed with the Bonferroni t-test. An extremely significant difference in silicone oil coverage was found

(Figures 9,A and B). This analysis was not done on the

among these groups (P

Figure 8. (Apple) Human lens: Gross photograph showing adherence of red dyed silicone oil.

<

.001). Assuming that the hu-

B

A

Figure 9. Silicone IOL coated with liquid perfluorocarbon. A: Gross photograph. B: Image analysis, mean percentage coating: 2.84%.

J CATARACT REFRACT SURG-VOL 23, MAY 1997

541

SILICONE OIL ADHESION TO 1015

w

0

zw a:: w

:r:

80

70

5 w z

80

0

40

u; ~ zw

30

C

:;

centage silicone oil adherence to optics of seven 10L biomaterials and the lens capsule.

90

Q

4( ...I

Figure 10. Comparison of mean per-

100

50

20

0

10

Go

0

a:: w

LENS BIOMATERIAL

man crystalline lens capsule represents the most physiologic structure, it was selected as a control to which the non-silicone IOL biomaterials were compared. Significantly greater silicone oil coverage was found on the MemoryLens and acrylic lenses (P < .01). No significant differences in silicone oil coverage were found

between heparin-surface-modified, fluorine-treated, hydrogel, and regular PMMA lenses and the lens capsule (Table 2). The power of the performed test was 0.944, indicating high validity of the results.

Discussion The interaction of silicone oil and silicone IOLs is a newly recognized clinical complication in vitreoretinal surgery and can result in visual disturbances, including decreased visual acuity and visual aberrations. We doc-

Air

umented three such cases in a previous reportl and rec-

Partial wetting

NOD-wetting

Biomaterial A

Contact Angle a=

00

- CQI11lIete

= > 00 < 900 -

Contact Angle a Contact Angle 9

=>

900

\\Elting

partial

v..etting

- noll\Wtting

B

Figure 11. (Apple) Contact angle and dispersive energy. A: Schematic diagram showing the concept of small versus large contact angle. B: A contact angle 90· connotes a nonwettable surface.

542

ommended that the use of silicone lenses 'in eyes with silicone oil or in the rare individuals at increased risk for requiring a vitreoretinal procedure be reconsidered. Indeed, the package information insert that accompanies AdatoSil 5000™ silicone oil marketed by Chiron Vision specifically contraindicates use of the oil in the presence of a silicone IOL because of the potential for the oil to chemically interact with silicone elastomers. Our data demonstrate that the interaction of silicone oil is greatest with silicone IOLs and that exposure of silicone IOLs to a hydrophilic, aqueous environment does not significantly reduce this interaction. Although each of the other IOL biomaterials showed significantly less coverage and adhesion of silicone oil, none of them was fully immune to this complication. Thus, the potential for visually significant silicone oil adhesion to other IOL materials may exist. With this in mind, it may be instructive to compare the results obtained with the

J CATARACT REFRACT SURG-VOL 23. MAY 1997

SILICONE OIL ADHESION TO IOLS

10Ls to the data obtained on the crystalline lens capsule. Our data showed an average of 10.7% coverage of the crystalline lens capsule by silicone oil. The adherence was not strong and the oil was easily removed with injected viscoelastic. Silicone oil is known to cause several significant complications that can affect vision. 5- 13 Despite its many years of use and study in Europe, silicone oil has not been reported to cause clinically significant adherence to the crystalline lens. This correlates with our finding of minimal in vitro interaction of silicone oil with the crystalline lens capsule, which would not be expected to cause a clinically significant interaction. It therefore seems reasonable to conclude that 10Ls made ofbiomaterials showing similarly low levels of silicone oil coverage would be less likely to result in a clinically significant interaction. It may be advantageous for an implanting surgeon to choose an 10L made of such a biomaterial. Of those we examined, heparin-surface-modified, fluorine-treated, hydrogel, and regular PMMA lenses showed no significant difference in silicone oil coverage compared with the crystalline lens capsule. Both the acrylic lens and the MemoryLens showed somewhat greater coverage. The physicochemical basis of the interaction of silicone oil with 10L biomaterials is incompletely understood. 14 - 18 Analysis of the surface chemistry and contact angle measurements of several 10L biomaterials has yielded data on some of the polar and nonpolar components of surface energy that affect their wettability; that is, the degree to which various liquids will spread across and coat their surface (Figures 11 ,A and B) (Cunanan, personal communication). Polar forces include both acidic and basic components, whereas nonpolar (dispersive) forces include Van der Waal's and · . wrces. c 14-18 London d IsperSlve Soft materials such as silicone have the ability to modify their surface energy in response to their environment. 19 When placed in an aqueous medium, normally hydrophobic silicone elastomers can reconfigure their surface energy to become more hydrophilic. Our data showed that silicone lenses that were first placed in normal saline had reduced silicone oil coverage, but this difference was not statistically significant compared with the coverage of silicone lenses in the dry state placed directly into silicone oil.

Silicone oil coverage of an 10L biomaterial appears to be influenced by the dispersive component of the 10L's surface energy. In our study, materials with low dispersive energy showed low levels of oil coverage, while materials with higher dispersive energy showed greater oil coverage. More extensive studies are needed to fully characterize all the surface energy components of various 10Ls, including specific determinations of the London and Van der Waal's components of dispersive energy. Relating this information to observed differences in silicone oil coverage would result in a greater understanding of the interaction of silicone oil with various 10L biomaterials. The data generated from such studies would assist in developing surface coatings for silicone lenses that would minimize their interaction with silicone oil.

References 1. Apple OJ, Federman JL, Krolicki TJ, et al. Irreversible

2.

3.

4.

5.

6.

7.

8.

9. 10.

11.

silicone oil adhesion to silicone intraocular lenses; a clinicopathologic analysis (discussion by T Aaberg Sr). Ophthalmology 1996; 103: 1555-1562 Batterbury M, Wong 0, Williams R, Bates R. The adherence of silicone oil to standard and heparin-coated PMMA intraocular lenses. Eye 1994; 8:547-549 Kusaka S, Kodama T, Ohashi Y. Condensation of silicone oil on the posterior surface of a silicone intraocular lens during vitrectomy. Am J Ophthalmol 1996; 121: 574-575 Stolba U, Binder S, Velikay M, Wedrich A. Intraocular silicone lenses in silicone oil: an experimental study. GraefesArch Clin Exp Ophthalmol1996; 234:55-57 Punnonen E, Laatikainen L, Ruusuvaara P, Setrua K. Silicone oil in retinal detachment surgery; results and complications. Acta Ophthalmol1989; 67:30-36 Hutton WL, Azen SP, Blumenkranz MS, et al. The effects of silicone oil removal; silicone study report 6. Arch Ophthalmol1994; 112:778-785 Fisk MJ, Cairns JO. Silicone oil insertion; a review of 127 consecutive cases. Aust NZ J Ophthalmol 1995; 23: 25-32 Maberley AL, Antworth MV. The use of silicone oil in vitreoretinal surgery. Can J Ophthalmol1989; 24:265268 Franks WA, Leaver PK. Removal of silicone oil-rewards and penalties. Eye 1991; 5:333-337 Riedel KG, Gabel VP, Neubauer L, et al. Intravitreal silicone oil injection: complications and treatment of 415 consecutive patients. Graefes Arch Clin Exp Ophthalmol1990; 228:19-23 Yeo JH, Glaser BM, Michels RG. Silicone oil in the treat-

J CATARACT REFRACT SURG-VOL 23, MAY 1997

543

SILICONE OIL ADHESION TO IOLS

12.

13.

14. 15.

16.

17.

ment of complicated retinal detachments. Ophthalmology 1987; 94:1109-1113 McCuen BWII, deJuan EJr, Landers MB III, Machemer R. Silicone oil in vitreoretinal surgery part 2: results and complications. Retina 1985; 5:198-205 Federman JL, Schubert HD. Complications associated with the use of silicone oil in 150 eyes after retina-vitreous surgery. Ophthalmology 1988; 95:870-876 Andrade JD Interfacial phenomena and biomaterials. Med Instrum 1973; 7:110-119 Baier RE. Surface properties influencing biological adhesion. In: Manly RS, ed, Adhesion in Biological Systems. New York, NY, Academic Press, 1970; 15-48 Hainsworth DP, Chen SN, Cox TA, Jaffe GJ. Condensation on polymethylmethacrylate, acrylic polymer, and silicone intraocular lenses after fluid-air exchange in rabbits. Ophthalmology 1996; 103:1410-1418 Lydon MJ, Minett TW, Tighe BJ. Cellular interactions with synthetic polymer surfaces in culture. Biomaterials

1985; 6:396-402 18. Mateo NB, Ratner BD. Relating the surface properties of

544

intraocular lens materials to endothelial cell adhesion damage. Invest Ophthalmol Vis Sci 1989; 30:853-860 19. Cunanan CM, Tarbaux NM, Knight PM. Surface properties of intraocular lens materials and their influence on in vitro cell adhesion. J Cataract Refract Surg 1991; 17:

767-773 From the Center for Research on Ocular Therapeutics and Biodevim, Department ofOphthalmology, Storm Eye Institute, Medical University ofSouth Carolina, Charleston, USA. Presented in part at the American Academy of Ophthalmology annual meeting, Atlanta, Georgia, USA, October 1995, and the Symposium on Cataract, IOL and Refractive Surgery, Seattle, Washington, USA, June 1996 Supported in part by an unrestricted grant from Research to Prevent Blindness, Inc, New York, New York. Richard Campbell and Thomas Piness, Robert Bosch Corporation, North Charleston, South Carolina, USA, assisted with the scanning electron microscopy. None ofthe authors has a proprietary interest in any material used in this study.

J CATARACT REFRACT SURG-VOL 23. MAY 1997