Postoperative membranous proliferation outgrowth onto the anterior surface of hydrophobic acrylic intraocular lenses

State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou 510060, China Guangdong Province Hospital for Occupational Disease Prevention and Treatment; Guangdong Provincial Key Laboratory of Occupational Disease Prevention and Treatment, Guangzhou, Guangdong 510300, China Department of Toxicology, School of Public Health and Tropical Medicine, Southern Medical University, 1023 South Shatai Road, Guangzhou 510515, China The first four authors contributed equally to this publication.


Introduction
Capsular opacification, or after-cataract, is the major complication of cataract surgery.It occurs due to migration, proliferation, and differentiation of residual lens epithelial cells (LECs).These LECs attempt to undergo normal differentiation or epithelial mesenchymal transition (EMT), forming heterogeneous groups of cells in the anterior capsule (anterior capsule opacification [ACO]) and posterior capsule (posterior capsule opacification [PCO]) [1] .
The purpose of this study was to identify the risk factors

RESEARCH ARTICLE
for MP on the IOL surface.We found that some LECs can proliferate and migrate from the anterior capsular margin onto the intraocular lens (IOL) to form a membranous sheet with an accumulation of extracellular matrix.This membranous proliferation(MP)may be related to the ACO and capsule contraction syndrome [2] .
We sought to elucidate the relationship between MP and ACO, and to determine which procedures can reduce ACO and capsule contraction syndrome simultaneously.

Patients and methods
This prospective study comprised 145 eyes from 135 consecutive patients who had cataract surgery at Zhongshan Ophthalmic Center from January 1 to August 1, 2014.Patients with ocular disease other than age-related cataract were excluded.Ocular axial length ranged from 22 mm to 26 mm.All studies of the patients were performed with the approval of the center's hospital ethics committee and conducted in accordance with the Declaration of Helsinki.Informed consent was obtained from all patients.All surgery was performed by the same experienced surgeon (Yizhi Liu), using a standard quick-chop technique [3-5]   .Phacoemulsification was performed with an Infiniti phacoemulsifier (Alcon Laboratories, Texas, USA).The surgical technique included a clear corneal incision in the temporal quadrant, a 5.5 mm continuous curvilinear capsulorhexis (CCC), phacoemulsification and aspiration, and in-the-bag IOL (SA60AT, Alcon, USA) implantation.Patients who had intraoperative complications, such as posterior capsule rupture, were excluded from the study.Eyes in which the margin of the anterior capsule did not completely cover the rim of the IOL optic because of excessive diameter or deviation of the CCC were also excluded.
Postoperatively, all patients received the same antibiotic eye drops for 1 month.One week and one month postoperatively, the pupil was dilated with tropicamide to allow examination of the IOL optic by slit-lamp microscopy, retroillumination, and diffuse illumination to document the morphology and extent of the cellular reaction following implantation.If MP was present, we re-examined the patient monthly to monitor the condition.The extent of proliferation was assessed using a clock-face representation of the IOL optic to identify the location of the MP.Whereas other studies have divided the 12 clock hours into four consecutive quadrants (Figure 1A) [6][7] , we divided them into four quadrants according to the position of the incision (Figure 1B).Differences in quadrants were evaluated by variance analysis (ANOVA).A level of P <0.05 was considered statistically significant.
Basic descriptive statistics were calculated for all data gathered and are reported as mean ± SD or n (%), as appropriate.Univariate logistic regressions were applied to those parameters that might be associated with the development of an MP.The parameters tested included age, gender, right vs left eye, preoperative visual acuity, ultrasound total time, ultrasound total equivalent power, postoperative visual acuity, ocular axial length, diopter power of the IOL, and grade of anterior vitreous opacity.Risk factors with P-values <0.15 were further evaluated with multiple logistic regression analyses.All tests were two-tailed and a P-value of <0.05 was considered statistically significant.
According to the degree of vitreous opacities proposed by Koga et al [8] , the degree of anterior vitreous opacity was classified as: Grade 1, mild, slight vitreous opacities ( Figure 2A); or Grade 2, moderate or severe, broad and dense vitreous opacities (Figure 2B).No intraoperative complications occurred.A total of 24 eyes were excluded from the study because the rim of the IOL optic was not completely covered by the anterior capsule margin and the axial length was abnormal.The frequency and characteristics of MP on the IOL optic were evaluated in the remaining 121 eyes.Table 1 shows the patient characteristics and outcome features.There was no statistical difference in MP between 1 week and 1 month (P>0.05).MP was detected in 46 eyes (38.02%) at 1 week.The membrane occurred as a transparent monolayer originating from the edge of the capsulorhexis.In all eyes except one, LEC growth was limited to the margin of the capsulorhexis and did not regress to the center.In one eye, the monolayer covered the entire anterior surface of the IOL at 1 week (Figure 2C), but disappeared at 1 month.In all cases, the anterior capsule was in contact with the IOL optic.This supports findings in previous studies describing the importance of the relationship of cell proliferation onto the optic [9] .

Correlation Factors
We did not find any correlation between MP and gender, right vs left eye，ultrasound total time, ultrasound total equivalent power, or postoperative visual acuity.However, univariate analysis did identify an inverse relationship between MP and age, preoperative visual acuity, axial length, diopter power of the IOL, and grade of anterior vitreous opacity.Multivariate logistic regression analysis was used to re-examine those characteristics identified in the univariate analyses for which P<0.15.This stepwise selection confirmed the relevance of age, axial length, and anterior segment vitreous opacity.The incidence of membranous outgrowth was greater in pseudophakic eyes characterized by shorter axial length, older age, and more serious anterior segment vitreous opacity (Table 2).

Features
Using the new method of dividing the lens surface according to the location of the incision, we found that the inferior (35.11%) and nasal (33.51%) quadrants were more likely to exhibit MP than the superior (18.09%) and temporal quadrants (13.30%) at 1 week.(See Table 3 and Figure 3.) Total area of MP did not substantially exceed that encompassed by a single quadrant (56.52% [Table 1 and Figure 4]).
We combined retroillumination and diffuse illumination with slit-lamp microscopy (Figure 5A, B, C) to analyze the morphology of MP, because the strong reflex from the IOL surface sometimes did not allow visualization of certain areas.Using the three methods, we could combine the three pictures and better visualize the total optic.The photographs in Figure 5 A, B, C are from the 1-month examination of a patient who had MP coveringal most three quadrants.Slit-lamp examination and microscopy showed that the MP was continuous with the capsular opening and concentric to the center of the optic, and it appeared to consist of elongated cells extending from the anterior capsular margin.The dendritic fan-shaped MP displayed an apical portion directed toward the central optics (Figure 5A).

AcrySof SA60AT
All of the patients received the recently introduced foldable hydrophobic acrylic IOL (AcrySof SA60AT, Alcon,USA), which is now commonly used.This IOL is a single-piece, square-edged acrylic with a 6.0 mm optic diameter and no haptic angulation.The SA60AT material is a co-polymer of 2-phenylethyl acrylate and 2-phenylethyl methacrylate with a water content of 0.3% [10] .
It has been reported that with a poly (methyl methacrylate [PMMA]) IOL, MP occurs in 40% to 80% of patients [11][12] .In our study, 46 patients (38.02%) displayed MP, which extended from the anterior capsule margin onto the IOL optic soon after cataract surgery.Kurosaka and Kato [13] found that the incidence and extent of MP were significantly greater on acrylic IOLs than on PMMA and silicone IOLs, due to improved biocompatibility of the acrylic IOL.Interestingly, Shah et al found that the incidence of ACO is lowest with acrylic lenses and highest with plate haptic silicone IOLs [1] .
We speculate that several factors may account for the conflicting findings mentioned above.First, cells forming ACO may derive from the germinal zone, whereas cells forming MP derive from the rim of the capsulorhexis.Second, all of these cells come from the rim of the capsulorhexis and the total number of cells is relatively constant.Cells that have more potential for adhesion and remain in the capsule will cause ACO, and those having more potential for migration and outgrowth onto the IOL surface will result in MP (Figure 5F).Third, MP cannot be ruled out as an index of lens epithelial cell proliferative activity [14] .Among those proteins related to the extracellular matrix, fibronectin promotes maximal migration of LECs, whereas type IV collagen promotes maximal adhesion of LECs in vitro; fibronectin is present in the aqueous humor of patients with cataracts [15] .

Correlation Factors
In contrast to our results, Lenis and Philipson did not observe any correlation between MP and other factors, such as age, capsulorhexis size, or presence of pseudoexfoliation [11]   .Reported risk factors for contraction of the anterior capsule opening include diabetes mellitus [16] , uveitis [17] , old age, and instability of the blood-aqueous barrier [18] .Risk factors for PCO include young age [19] and pseudoexfoliation [20] .Pathologically, contraction of the anterior capsule opening and PCO are considered to result from similar mechanisms; however, the risk factors for the two disorders are not identical.In particular, they have opposite relationships to age.If the cells derived from capsule contraction syndrome were those with greater potential for migration, whereas cells from PCO were those with greater potential for adhesion, the two cell types would not have opposite relationships with age.This can account for the different relation between age and contraction of the anterior capsule opening and age and PCO.
Older patients may have different levels of growth factors in the anterior chamber compared to younger patients, and this may be related to the potential of LECs for adhesion and migration.We had collected some samples of aqueous humor to elucidate which factors influence cell activity and the ratio of migration factors to adhesion factors, which could simultaneously lessen ACO and capsule contraction syndrome.
In patients who had shorter ocular axes, shallower anterior chamber depth, narrower space in the anterior ocular segment, and higher vitreous pressure, the operation was difficult to perform.The postoperative inflammatory reaction was severe and may account for the shorter ocular axial length and serious MP observed.Some studies have found that ACO [21] and MP [7] were related to anterior chamber laser flare; proliferation at the anterior interface occurs on the portion of the implant surface exposed to the aqueous humor.We propose that both the anterior and the posterior environments influence lens epithelial cell proliferation on the anterior implant surface.In our study, we found that vitreous opacity was correlated with proliferation; inflammatory factors from the vitreous humor may have contributed to the process of membrane formation, because some cells and proteins could move freely between the anterior and posterior chambers with the aqueous humor.
Furthermore, we propose that human lens epithelial cell proliferation and migration in vivo are initiated by loss of contact inhibition as a result of surgical trauma.We also propose that humoral factors are released into the aqueous humor as a result of postoperative breakdown of the blood-aqueous barrier [11] .
Our results with regard to ultrasound total time and ultrasound total equivalent power were similar to those of Kruger et al [9] .Ultrasound time and power influence intraoperative LEC damage.In addition, the prolonged operating time resulting from relatively large quantities of fluids used to irrigate the anterior chamber might influence intraoperative LEC damage, while more time and power may also promote LEC.
We speculate that the location of the incision was correlated with MP, which has not been previously reported.Obviously, the location of the incision affects the operative procedure.Both the CCC and the aspiration of the cortex is limited in a small space that is difficult to reach, which may influence the removal of the LECs and the potential proliferative ability of LECs.Further study is needed to elucidate these effects.Moreover, the influence of gravity may increase the density of certain factors in some quadrants.If the concentration of some factors related to LEC migration in the inferior quadrant was denser than in the other quadrants, the MP may be more serious in this quadrant.

Morphology
Lens epithelium generates an endogenous physiological electric field (EF) [22] , in which LECs proliferate, migrate, re-orient, elongate and differentiate into lens fibers.LECs migrate in one of two directions in vivo.Either they migrate toward the equator and elongate and trans-differentiate to become lens fiber cells, or they migrate in the opposite direction, toward the front of the lens, to compensate for ongoing apoptosis of LECs as the epithelium turns over.This EF may be related to the dendritic nature of fan-shaped membranous growth, which exhibits an apical portion directed toward the central optics (Figure 5A).
Usually, over continued follow-up, MP disappeared within a few months or retracted as a thinner membrane close to the capsulorhexis (Figure 5D).However, there were some exceptions, as described below.One patient exhibited a gap between the MP and the margin of the capsulorhexis at 1 month (Figure 5E), but this gap had completely disappeared at 2 months.
Another patient displayed a substance likely to be composed of cells extending throughout the entire IOL surface at 1 week (Figure 2 C), but this proliferation disappeared at 1 month.This may have occurred because LECs were not immediately removed by aspiration when they emerged on the surface.The LECs were dispersed throughout the viscoelastic material and remained there until the viscoelastic substance was removed at the end of surgery.It is possible that the LECs adhered to the capsule or IOL surface, proliferated, and migrated.These two cases suggest that certain factors of the aqueous may affect the metabolism of residual LECs on the surface of the IOL.
A third patient displayed differences between the left and right eyes, with serious MP occurring in one eye (Figure 5A,  B, C) and none in the other.Individual variations in the intraocular conditions that render the surrounding conditions sufficient for LEC proliferation cannot be ruled out.This could explain the differences in proliferative capacity between individual patients or different eyes.

Possible Etiologic Mechanisms
In general, it is believed that LECs from the equator of the capsular bag play a minor role in the development of anterior cell migration and capsule opacification [23] .LECs that proliferate onto the anterior IOL surface are thought to originate from the anterior capsule close to the capsulorhexis.We tend to share this opinion, based on our clinical observation of 121 cases and on various observations reported in the literature.A number of etiologic mechanisms may contribute to MP, as discussed below.
Residual LECs proliferate and transform metaplastically into myofibroblasts after cataract removal.Proliferating LECs migrate from the anterior capsule margin onto the anterior optic surface, accompanied by an accumulation of extracellular matrix.It is possible that an increased inflammatory response to surgery in some patients promotes formation of this fibrous membrane.
Fibrous metaplasia of LECs usually occurs between the IOL optic and the inner surface of the anterior capsule [24] .
The fibrous membrane is present near the anterior capsule, and its contraction decreases the anterior capsule opening.These findings suggest that lens epithelial cell outgrowth from the capsule margin onto the IOL may be related to capsule contraction syndrome [25] .
Multiple extracellular molecules may be involved in the in vivo adhesion and migration of lens epithelial cells.This suggests that multiple independent mechanisms can be involved in this response or that multiple interdependent steps are involved in either the adhesion or migration of these cells in response to extracellular matrix proteins.
It has been postulated that fibronectin of serum origin resulting from postoperative damage to the blood-aqueous barrier could play an important role in lens epithelial cell proliferation and migration [26] .Notably, fibronectin is not found in significant quantities in adult human lens capsule [27]   , but fibronectin immunoreactivity has been demonstrated in the acellular proteinaceous membrane on explanted IOLs.
In dose-response experiments, fibronectin promoted maximal migration at lower concentrations in comparison with type IV collagen and laminin [28] .
Ultimately, a sound understanding of the molecular mechanisms directing lens epithelial cell adhesion and migration should provide direction for the development of novel biopharmaceutical agents to prevent vision-disturbing complications after cataract surgery.Further studies are necessary to prove that the membrane reported in this study is the LEC sheet and to determine which factors in the aqueous humor are related to MP.

Figure 1 .
Figure 1.The extent of MP was assessed using a clock-face representation of the IOL optic as the number of area of clock hours in which membranous proliferation could be identified.Figure A shows the clock hour divisions commonly used in other studies.Instead, we divided the clock hours into 4 quadrants according to the position of the incision (Figure B).In all cases, the incision was made in the temporal region (Quadrant 1.).Quadrant 1: 1-4 o'clock.Quadrant 2: 4-7 o'clock.Quadrant 3: 7-10 o'clock.Quadrant 4: 10-1 o'clock.
Figure 1.The extent of MP was assessed using a clock-face representation of the IOL optic as the number of area of clock hours in which membranous proliferation could be identified.Figure A shows the clock hour divisions commonly used in other studies.Instead, we divided the clock hours into 4 quadrants according to the position of the incision (Figure B).In all cases, the incision was made in the temporal region (Quadrant 1.).Quadrant 1: 1-4 o'clock.Quadrant 2: 4-7 o'clock.Quadrant 3: 7-10 o'clock.Quadrant 4: 10-1 o'clock.

Figure 2 .
Figure 2. Anterior vitreous opacity.A: Grade 1, mild, slight vitreous opacities; B: Grade 2, moderate or severe, broad and dense vitreous opacities.C: A substance likely to be cells covered the entire anterior surface of the IOL at 1 week.

Figure 3 .
Figure 3.In 46 eyes with MP, the number of hours affected in each quadrant of the optic was determined.I.e., if an eye had MP in the area of 2-4 o'clock and in the area of 4-5o'clock, that optic was recorded as having 2 hours in quadrant 1 and 1 hour in quadrant 2. Clock hours affected were 25, 66, 63, and 33 in quadrants 1-4, respectively.Quadrants: 1= temporal; 2=inferior; 3= nasal; 4=superior.

Figure 5 .
Figure 5.The photographs in Figure 5 A, B, C are from the 1-month examination of a patient who had MP covering almost three quadrants.We combined retroillumination and diffuse illumination with slit-lamp microscopy (A, B, C) to analyze the morphology of MP.A: Slit-lamp microscopy shows MP continuous with capsular opening; arrow points to the center of the optic.B: Diffuse illumination shows MP.C: Retroillumination shows MP.D: Linear margin at 1 month.MP retracted and became a thinner membrane close to the capsulorhexis.E: Discontinuity in the process of regression.The retraction was usually contiguous (as in Figure D), but one patient exhibited a gap between the MP and the margin of the capsulorhexis at 1 month (E).F: Membranous proliferation between anterior capsule and the IOL and on the surface of the IOL.

Table 2 . Summary of relationship variables and presence of MP in pseudophakic eyes
*Only variables with a P value of less than 0.15 in the univariate analysis were included in the multivariate model.