In Vivo Observation of Lens Regeneration in Rat Using Ultra-Long Scan Depth Optical Coherence Tomography

Abstract

Purpose: To evaluate morphologic changes of lens regeneration in rats in vivo after extracapsular lens extraction (ECLE) by ultra-long scan depth optical coherence tomography (UL-OCT).

Methods: A total of 42 Sprague-Dawley rats were used in this study. We performed ECLE on the right eyes of animals in the surgery group (n = 34). Biomicroscopy and UL-OCT scans were carried out for the surgery group immediately (within 1 hour postoperatively) and at days 1 and 3, weeks 1 and 2, and months 1, 2, and 3 postoperatively. After in vivo examination, three animals of the surgery group were euthanized at each time point for histology study, while the other 10 animals were examined continuously at those time points. The regenerated lens was evaluated in OCT images at 2 and 3 months postoperatively. The control group consisted of eight untreated rats that had OCT examination at the age of 5 months.

Results: Lens regeneration could be observed from 2 weeks postoperatively. Regeneration was mainly at the peripheral capsular bag in the first month and central region thereafter. The average thickness of regenerated lenses was 2222 ± 309 and 2324 ± 352 μm at 2 and 3 months, respectively. Regeneration was faster in the first 2 months and slowed down thereafter. Although anterior capsule opening and posterior capsule adhesion and wrinkling existed, the regenerated lens still could form a relatively regular shape, however, the size was much smaller than that of the normal lenses from rats with the same age.

Conclusions: Ultra-long OCT provides longitudinal data of the process of lens regeneration on a single individual rat in vivo, which may allow one to follow and compare the lens regenerative process under different interventions or therapy after ECLE in rats.

Organisms have the ability to regenerate body parts or organ in response to injury. Studies on regeneration of body parts or organ in invertebrates and amphibian can be found over two centuries ago. Colucci¹ and Wolff² are the first two to observe lens regeneration in adult newts independently. Newt lens can regenerate from the dorsal root of the iris. This process is called transdifferentiation or Wolffian regeneration. The mammalian lens also exhibits some level of regeneration ability through a different way, in which a new lens arises from residual lens epithelial cells (LECs) and requires a relatively intact lens capsular bag. The most recent work by Lin et al.,³ which successfully demonstrated lens regeneration taking place in the eyes of human infants, has ignited widespread interest. On the one hand, studies of the mechanism of lens regeneration can help to successfully rebuild a human lens after lens extraction or damage. On the other hand, the prevention of posterior capsular opacification (PCO) forming after cataract surgery can also be achieved by promoting lens regeneration.⁴ 

However, the actual morphologic changes in vivo based on continuous high-definition observation of lens regeneration process are still unknown. The conventional way to study lens regeneration is through morphology observation by histopathology or slit-lamp biomicroscopy, the former one is invasive and takes a long time for a series of procedures, while the latter one provides limited information by assessment of the lens in a cross-sectional manner. Furthermore, previous works show that lens thickness in the central plane measured by A-scan biometry and Scheimpflug does not give the resolution that can potentially be obtained by OCT imaging.⁵ Optical coherence tomography (OCT) is an advanced optical imaging modality, which can obtain high-resolution cross-sectional imaging of tissue in vivo, with or without transparency, noninvasively and at a high speed. It has been used in examination and analysis of the retina,^6,7 nerve fiber layer^8,9, cornea,^10,11 and other structures in anterior segments.^10,11 In lens study, it has also been used to assess cataract¹² and PCO formation^13–19 in human eyes. Since OCT is an ideal tool to observe the structural changes of tissues longitudinally, it has been applied in contact lens edge fitting^20,21 and tear film assessment.^22–24 Recently, a novel OCT system (ultra-long scan depth [UL]-OCT) with prolonged scan depth of 7.2 mm has been developed, which can image the entire ocular surface and overall lens in the human eye.²⁰ It offers the possibility of whole eyeball imaging in rodents.²⁵ 

This study aims to assess lens regeneration after extracapsular lens extraction (ECLE) in vivo by UL-OCT, and explore the pattern of morphologic changes of the process by continuous observation. 

Methods

Animals

A total of 42 Sprague-Dawley rats (age 2 months and weighing 200–250 g) with clear lenses were used in this study; 34 of them were placed in the surgery group and the other 8 animals were placed in the control group. All animal procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The Institutional Review Board for Animal Research of Wenzhou Medical University approved this study. 

Optical Coherence Tomography

A custom-built UL-OCT prototype (see Supplementary Fig. S1) was used to assess lens regeneration in vivo after surgery by taking an image of the horizontal meridian of the eyeball. The detail specifications of the facility were described in the previous study.²⁰ In brief, this prototype OCT was based on the technology of spectral-domain OCT, with modified transmission grating and a line scan charge-coupled device (CCD) camera (Aviiva SM2 CL 2010, 2048 pixels; Atmel, San Jose, CA, USA), which has a wavelength of 840 nm, a scanning depth of 7.392 mm in air, and an axial resolution of 6 μm. In this study, the maximum scan depth was applied to acquire the image of the entire eyeball. 

The optical coherence tomography probe with video monitor was mounted in parallel on a modified slit lamp base. Both the x- and y- axial real-time OCT images acquired by the scan CCD camera and the real-time video acquired by the video camera were displayed on a computer screen to help to aim. An x-y crosshair was applied in real-time video to indicate the scan position. 

Surgical Procedure

Detail surgical procedure was described in the previous study²⁶ (see Supplementary Fig. S2). We performed ECLE on right eyes of 34 rats, and one surgeon (KZ) performed all procedures. Animals were anesthetized with pentobarbital sodium 40 mg/kg by intraperitoneal injection. We applied 0.05% proparacaine hydrochloride for topical anesthesia. The pupil was dilated adequately by repeated instillations of 1.0% tropicamide. A corneal incision was made with a 15° disposable stab knife. After the injection of 1% sodium hyaluronate into the anterior chamber, the corneal incision was extended from 120° to 150° with fine Vannas scissors. An anterior curvilinear continuous capsulorrhexis (CCC) of approximately 3 mm in diameter was created by capsulorrhexis forceps and followed by hydrodissection and lens removal. Then anterior chamber and capsular bag were cleaned by saline solution. Finally, the corneal incision was sutured with interrupted 10-0 nylon sutures, and topical chloramphenicol was used at the end of surgery. 

All treated eyes were administrated with 1.0% tropicamide and chloramphenicol twice daily for 3 days or until rats were killed post surgery. 

Examination

Animals were anesthetized with intraperitoneal injection of pentobarbital sodium at a dose of 40 mg/kg,; their pupils were dilated with 1.0% tropicamide. For the 34 animals in the surgery group, biomicroscopy and OCT examinations were administrated immediately (within 1 hour postoperatively) and at days 1 and 3, weeks 1 and 2, and months 1, 2, and 3 after surgery. Ten of 34 animals were examined and photographed at each time point until 3 months after surgery. For the remaining 24 animals, 3 of them were euthanized with pentobarbital sodium by intraperitoneal injection at a lethal dose of 100 mg/kg at each time point; their treated eyes were enucleated and processed for histologic study. Considering the potential effects of surgical stimulation on the contralateral untreated eye,²⁷ eight animals in the control group were left without surgery. The control group had OCT examination only when they reached the age of 5 months, which was the same age as the surgery group at 3 months postoperatively. The flow of this study is shown in Figure 1. 

Figure 1

 

Flow diagram of experimental procedures.

Optical Coherence Tomography Examination and Measurement

Animals were placed on a custom-built stage (see Supplementary Fig. S1) with their cornea aligned in the center and the x-y crosshair aiming through the real-time video. Adjusting the probe to make the ocular surface aligned with the horizontal line in both x- and y-axis of OCT imaging. Images on the x-axis were recorded manually with an operator as soon as corneal apex reflection was detected in both x- and y-axis imaging. Optical coherence tomography imaging of each individual rat at each time point was acquired twice with the probe realigned each time and proceeded with further measurements and statistical analysis. 

Raw OCT images were imported in custom software and calibrated for measurement. Regenerated lens thickness was defined as the maximum distance from the anterior surface to posterior surface of lens capsule in the z dimension (Fig. 2). The regenerated lens thickness was calculated as the average thickness of two measurements at each time point. Since regenerative material filled the capsular bag at 2 months postoperatively, 2 and 3 months postoperatively were selected as the comparison time points. While the thickness immediately after surgery was set as a baseline of regeneration thickness, the value was set to 0. The lens thickness in 5-month-old rats was also measured as the control. An optical coherence tomography image of a normal rat eye is illustrated in Supplementary Figure S3. If the animal's OCT image was not suitable for further analyses, this animal would be excluded (i.e., regenerated lens was too cloudy to clearly show the anterior or posterior surface clearly in OCT image). 

Figure 2

 

Measurement of regenerated lens thickness in OCT images (flipped vertically and cropped). The distance between the anterior and posterior surfaces represented the regeneration thickness.

The refractive index of the equivalent homogeneous normal lens and regenerated lens at an 840-nm wavelength could be approximated with a simplified Sellmeier formula²⁸:  

 

where n is the refractive index, λ is the wavelength, and A and B are coefficients that can be determined by fitting the equation to measured refractive indices at known wavelengths. In the present study, our data were based on Chaudhuri's²⁹ data. Therefore, the refractive index of a normal lens is 1.666.  

The refractive index of a regenerated lens was not referred to in the previous study. Considering the main component of a newly regenerated lens was elongated lens fibers, which was close to the lens cortex, we used the refractive index of a normal lens cortex instead, which is 1.411. 

Slit Lamp Biomicroscopy and Photography

Lens regeneration was evaluated by slit lamp biomicroscopy using direct illumination and retroillumination, then images were photographed by a single reflex digital camera (Nikon D5100; Nikon, Inc., Tokyo, Japan). 

Histopathology

After animals were euthanized with an overdose of pentobarbital sodium, eyes after surgery were enucleated under an operating microscope. After fixing eyes in 4% paraformaldehyde overnight, they were dehydrated in 30% sucrose solution for approximately 24 hours, and then embedded in optimal cutting temperature compound. After eyes were frozen, 12- to 14-μm sections were cut and stained with hematoxylin and eosin. 

Statistical Analysis

Longitudinal data and normal lens data were included for statistical analyses. Paired t-test was applied to determine the differences between the two consecutive measurements at the same time point. Independent sample t-tests were used to compare the mean thickness between the regenerated lens and normal lens 3 months after surgery. The mean thickness of comparison was initially calculated using Mauchly's test of sphericity to determine the subsequent statistical test method. If the data met the assumption of sphericity, ANOVA was used to determine the difference directly; if failed, the P value of the ANOVA was adjusted by the Greenhouse-Geisser method. Bonferroni test was used for the pairwise comparison. Analyses were performed with statistical software (SPSS 21.0; IBM Corp., Armonk, NY, USA). All statistics were 2-tailed; statistical significance was defined as P < 0.05. 

Results

All animals (n = 42), including the control group (n = 8), had slit lamp biomicroscopy and OCT examined; 24 out of 42 animals were used for histopathologic study. Among 10 animals under continuous OCT observation, seven had distinct OCT images with fairly regular shape for further measurements and analyses, and three were excluded because their regenerated material was too cloudy for imaging the whole lens or had a significantly irregular regenerated shape. Among these seven animals, we found 3 animals (42.9%) that were fairly translucent in the CCC region; the other 4 animals were not translucent in the CCC region but still could be imaged in OCT. 

Slit Lamp Biomicroscopy and OCT Findings

The postoperative slit lamp biomicroscopy and OCT images were shown in Figures 3 and 4. 

Figure 3

 

Optical coherence tomography images (A, D, F, H); biomicroscopy photography images with retroillumination (×25) (B, E, G, I); and direct illumination (×25) (C) of continuous observation on lens regeneration in rats from immediately to 1 week postoperatively. Optical coherence tomography images were flipped vertically. Immediately after surgery, in the OCT image, wave-like anterior and posterior capsules did not adhere together ([A] white arrow) and the CCC rim presented as a curve ([A] white arrowhead). Biomicroscopy photograph showed the anterior chamber was a little turbid, with exudate or residual viscoelastic agent (B, C). One day after surgery, the anterior chamber became clear, the anterior and posterior capsules adhered together, and the capsular bag was flat and clear (E), and the CCC rim was still clearly visible ([D] white arrowhead). Three days after surgery, the posterior capsule at the CCC region became thick in the OCT image ([F] between white arrowheads); in biomicroscopy photography the CCC rim shrank, and the posterior capsule folded at the CCC area (G). Seven days after surgery, the posterior capsule thickened, the posterior surface showed folds ([H] between white arrowheads) in OCT, and the CCC region decreased after the capsular bag shrank (I).

Figure 4

 

Optical coherence tomography images (A, D, G, L, P); biomicroscopy photography images with direct illumination (B, E, H, M, O); and retroillumination (C, F, I, J, K, N) of continuous observation on lens regeneration in rats from 2 weeks to 3 months postoperatively. Two weeks after surgery, the OCT image showed that the peripheral region thickened ([A] white arrowheads), which was in accordance with the biomicroscopy photograph ([B] black arrowheads); the posterior capsule at the CCC region was still folded (C). One month after surgery, the periphery thickened more remarkably than the central region ([D] white arrowheads) and the shape of the lens capsule formed a “dumbbell” in both the OCT (D) and the biomicroscopy photography image (E). The CCC region didn't thicken much and formed a “cup” (F). Two to 3 months after surgery, the thickness at the central region increased much faster; the shape appeared spheroid but with disorganized lens structure (G, L). The regenerated lens presented an intense inhomogeneous signal inside in the OCT images ([G, L] white arrows). Vacuoles were noticed mainly at the peripheral region ([I, N] black arrows; [J]), the CCC rim was clear ([K] black arrowheads), the matter at the CCC region did not extrude (H, M). Newly regenerated lens material was translucent with some haziness inside (B, E, H, M), not as transparent as a normal lens. In some cases, new regenerated lenses were opaque (O) and the matter at the CCC region extruded to the anterior chamber ([P] white arrow).

Immediately after surgery, both slit lamp and OCT observations showed the anterior chamber deepened and the anterior CCC rim laid against the posterior capsule (Figs. 3B, 3C). One day after ECLE, the CCC rim touched the posterior capsule and the posterior capsule presented smooth and clear under the slit lamp (Fig. 3E). The relative location of anterior and posterior capsules could be recognized more intuitively in the OCT image. The structures, presented as a smooth and flattened line with a curly CCC rim (Fig. 3D), did not adhere right after surgery (Fig. 3A), but at day 1 postoperatively. From day 3 to week 1 postoperatively, slit lamp biomicroscopy showed the CCC region had shrunk and the capsule contained wrinkles (Figs. 3G, 3I). The optical coherence tomography image showed that the capsule at the CCC region shrank and thickened at day 3 (Fig. 3F), and became smaller and thicker at week 1 (Fig. 3H). 

After 2 weeks, peripheral capsular bags thickened and remained transparent (Fig. 4A for OCT; Figs. 4B, 4C for slit lamp). Lens regeneration was faster at the peripheral region than the CCC region from 2 weeks to 1 month after surgery (Figs. 4E, 4F), and at 1 month, the capsular bag was in “dumbbell” shape in the OCT image (Fig. 4D). During 2 to 3 months, regeneration filled the CCC region. Although the anterior capsule was not intact (Fig. 4K), some regenerated lenses still formed a spheroid shape, with round smooth anterior surface and translucent material (OCT: Figs. 4G, 4L; slit lamp: Figs. 4H, 4I, 4M, 4N). In almost all surgery cases, they appeared with mild turbidity and had vacuoles located in the peripheral region (Fig. 4I, 4J). In optical coherence tomography, they presented as nonuniform high signal (Figs. 4G, 4L). For some irregular regenerated lenses, the protrusion at anterior surface and moderate to severe opacities were noticed (Figs. 4O, 4P). 

Histopathology and OCT Findings

In histopathology sections immediately after ECLE, a monolayer of LECs was observed at the anterior capsule and equator while anterior and posterior capsules had not adhered yet (data not showed). One day after surgery, the anterior and posterior capsules contacted and the capsular bag was filled by single layers of LECs. Little or no LECs were found at the posterior capsule of the CCC region, except the area near the CCC rim (Figs. 5B, 5C). Three days after surgery, LECs in the capsular bag and near the CCC rim proliferated into multiple layers (Figs. 5F, 5G). 

Figure 5

 

Optical coherence tomography images (A, D, E) and histopathology sections (B, C, F–J) of lens regeneration in rats from day 1 to day 7 postoperatively. Optical coherence tomography images were flipped vertically and cropped. One day after surgery, the anterior and posterior capsules contacted, the curly CCC rim could be identified in the OCT image (A) and in the histopathology sections (B), the capsular bag was filled by single layers of LECs. The posterior capsule at the CCC region presented little or no LECs, except the area near the CCC rim ([C] enlarged image of dashed box area in [B]). Three days after surgery, LECs in the capsular bag and near the CCC rim proliferated to multiple layers (F), double layers in the peripheral capsular bag ([G] enlarged image of dashed box area in [F]). From 3 to 7 days (OCT images: [D, E]; histopathology sections: [F–J]), proliferation and differentiation started at the peripheral region, the lens bow region was formed, and there were areas of LEC monolayer on the anterior capsule, as well as areas of hyperproliferation and abnormal LEC shape ([I] enlarged image of lower dashed box area in [H]). The posterior capsule near the CCC rim presented wrinkles ([J] enlarged image of upper dashed box area in [H]). Scattered fiber cell nucleus, Elschnig pearl-like structures, vacuoles, and breaks could be found inside the capsular bag and around the CCC rim. In the OCT image some structures could be identified as in the histopathology sections, like the curly CCC rim, a thickened posterior capsule, and a capsular bag (double-headed arrows, between [A, B, E, H]). Scale bars: 300 μm (B, F, H); 100 μm (C, I, J).

From day 7 to 30, LECs at the equator continued to proliferate, elongating or differentiating into lens fiber cells or lens fibers to fill the capsular bag and formed the Soemmerring's ring, while LECs at the center region had less differentiation. Thus, lens bow structure was formed (Figs. 5H–J, 6B–D, 6F, 6H–J). Between the months 2 and 3 postoperatively, adhesion between the anterior and posterior capsules at the CCC region departed and more lens fibers differentiated by LECs filled the center region. At 2 months, the spheroid-shaped capsular bag was filled with regenerative lens tissue, but had poor alignment of lens fibers inside, beside the equatorial zone (Figs. 6L–N, 7L, 7M). Aberrant distribution of lens fiber nucleus, as well as Elschnig pearl-like bodies, vacuoles, and breaks, could be seen inside (histopathology images in Figs. 6, 7L). 

Figure 6

 

Optical coherence tomography images (A, E, G, K) and histopathology sections (B–D, F, H–J, L–N) of lens regeneration in rats from 2 weeks to 2 months postoperatively. Optical coherence tomography images were flipped vertically and cropped. Two weeks to 1 month after surgery, peripheral region thickened obviously, and the posterior capsule at the CCC region showed wrinkles ([A, E] white arrowheads; [B, F] black arrowheads; [G] enlarged OCT images of white arrowhead area in [E]), and cells started migrating to the CCC region ([H, I] enlarged images of the right and middle dashed box areas in [F], respectively). The anterior capsule showed single layer cells ([C] enlarged image of left dashed box area in [B]; [J] enlarged image of left dashed box area in [F]), and the lens bow region was close to normal lens structure ([D], enlarged image of right dashed box area in [B]; [J] enlarged image of left dashed box area in [F]). At 2 months after surgery, the regenerated lens appears spheroid with disorganized lens structure (K, L), except in the equatorial zone ([M, N] enlarged images of left and right dashed box areas in [L], respectively). Inside new lens, some misalignment of fibers, breaks between fibers, aberrant distribution of lens fiber nucleus, as well as Elschnig pearl-like bodies and vacuoles could be seen (B–D, F, H–J, L–N). Scale bars: 500 μm (B, F, L); 50 μm (C, D, H–J, M, N).

Figure 7

 

Slit lamp, OCT, and histopathology image comparison between normal lens and regeneration lenses. Normal lens of 5-month-old rat (A–C, J, K). Lens regeneration with more regular shape and translucence after ECLE, but it still appeared hazy when compared with the normal lens (D–F, L, M). Lens regeneration with irregular shape and opacity (black arrowheads) (G–I). Histopathology of lens bow regions in regeneration and normal lens were shown in higher magnification (K, M). In optical coherence tomography images (C, F, I), the blue dotted line shows the anterior capsule, the red dotted line shows the posterior capsule. Scale bars: 500 μm (J, L); 50 μm (K, M).

Although the resolution of OCT images (Figs. 5, 6) was not able to identify a single cell as histopathology, some detailed structures were identified, including a curly CCC rim, a thickened peripheral capsule bag, vacuoles or turbidity, and the shape of a capsular bag. These OCT findings corresponded with histopathology findings, but not as clear as histopathology and slit lamp biomicroscopy. 

Figure 7 shows the slit lamp, OCT, and histopathologic images comparing normal lens and regeneration lens. A good regenerated lens could be seen, which was more regular in shape and translucent in material, but still more hazy than the normal one. The histopathology section shows that the bow region in the regenerated lens was similar to the normal one. 

OCT Measurements

There was no significant difference between two consecutive OCT measurements (all P > 0.05). Mean thicknesses of regenerated lens at 2 and 3 months after surgery were 2222 ± 309 and 2324 ± 352 μm (mean ± standard deviations), respectively. There were significant differences in regeneration thickness among the three time points (P < 0.001); there was a significant difference between baseline and 2 months (P < 0.001), but no significant difference was found between 2 and 3 months (P = 0.688). The thickness of regenerated lens increased faster within the first 2 months and slowed down thereafter. 

The mean thickness of normal lenses of 5-month-old rats was 3586 ± 43 μm. There was a significant difference between the thickness of normal lens and regenerated lens of 3 months after surgery (P < 0.001). The size of the regenerated lens was much smaller than the normal lens. 

Discussion

In previous studies, different methods for the evaluation of lens regeneration have been reported, mostly by histopathologic sections. Other methods include anterior and posterior segment photography, weighting,^30,31 protein analysis, and current measurement.³² All these ex vitro methods require killing the animals, and are not able to provide longitudinal data over the same subjects. There are in vivo ways, like slit lamp biomicroscope photography, digital image analysis for the lens regrowth area,^30,33 A-scan biometry for measuring the lens regrowth thickness.³⁴ However, they have limitations. Slit lamp biomicroscope photography is a conventional method for clinic examination; retroillumination images are useful for lens opacity and PCO evaluation.³⁵ Slit lamp provides intuitive observation of regenerated tissue, but without precise measurement. Digital image analysis is good for area assessment, but hardly provides in-depth information. As for the last approach, A-scan biometry gives comparable precision axial dimension measurement of lens thickness as OCT³⁶ and detects vacuoles and opacity in the lens, but it requires direct contact with the eye and local anesthesia, which increases the risk of infection. Expertise is also required to align the probe ideally in line with the visual axis; otherwise, it may result in inaccurate measurements. Alternatives such as optical A-scans including IOL Master and Lenstar can be adopted to avoid these disadvantages, nevertheless, they still can't produce a cross-section image of the regenerated lens. 

In the present study, we applied a novel noninvasive optical method, OCT, to trace and assess lens regeneration in the rat model, which allows cross-sectional imaging of the regenerative material in vivo at a high resolution. In the optical coherence tomography image, the shape of the regenerated lens—and structures like anterior and posterior capsules, CCC rim, capsular bag fold and thickening, some vacuoles or opacity inside regenerated lens—can be identified, which corresponds with histopathology sections and slit lamp biomicroscopy, but is not as clear as them. Optical coherence tomography can give biometry measures as precise and repeatable as other techniques. Repeated measurements on a single individual make the experiment more efficient and help to keep the variability low. Pentacam is another noninvasive modality to image the anterior segment; the previous study used Pentacam to measure lens thickness,^37–39 quantify PCO,⁴⁰ and a recent study used it to image the regenerated lens.³ Pentacam is characterized by a large depth of focus, which allows full cross-sections from the anterior cornea to the posterior lens in the Scheimpflug image. Quantitative information on the crystalline lens can be acquired as well, but cannot compete with the resolution of anterior segment OCT. 

Ultra-high resolution OCT (UHR-OCT) provides more detailed and high resolution images of the anterior segment for evaluating the human PCO,⁴¹ but is limited to a tissue scan depth of approximately 2 mm,^42,43 while the thickness of the entire regenerated lens at a later stage will reach more than 2.3 mm, which is beyond the scan range of UHR-OCT. Ultra-long OCT has the advantage over UHR-OCT because of its greater scanning depth. It also provided relatively high-definition observation and images in the present study. 

So far, all optical detection techniques have some major limitations. First, they are subject to image distortion due to the refraction of the intervening surfaces.^44,45 Correction of distortion is critical to obtain a reliable measurement. The refractive index of a regenerated lens was not referred to in the previous study. Although we used the lens cortex instead, the true refractive index is still needed to give a further correction. Second, the iris pigment blocks the light to prevent visualizing the structures behind, including the lens equator and zonules. Magnetic resonance imaging and ultrasound biomicroscopy show the potential value in whole natural or regenerated lens imaging. 

We use thickness as a scale of regeneration, since it is a useful factor to assess lens regeneration³⁴ and PCO.⁴⁶ In our study, lens regeneration proceeded faster at the first 2 months in rats, and the pace slowed down thereafter. Huang and Xie³¹ showed that the weight of rats' regenerated lenses increased mostly (0.013 g) in month 1, and only 0.0068 and 0.0025 g for months 2 and 3. Lois et al.³² showed that complete lens regeneration in rats occurred 8 weeks postoperatively. Our study is in accordance with theirs. However, further study is still needed to determine whether a regenerated lens could develop into the same size as a normal lens after a longer period. 

In our study, some cases showed the thicknesses of regenerated lenses decreased at month 3 after surgery, which may be due to structural rearrangement or remodeling of new lenses. In addition, more advanced and precise objective positioning system is needed, since poor fixation of animals under general anesthesia will affect the reliability and repeatability of the measurement. Optical coherence tomography is an advanced technology and we anticipate the custom-built prototype in our current study to undergo further development. At a later stage of regeneration, OCT images were found to have poor scan quality. Since this device was a custom-built OCT prototype using an 840-nm light source, the scan quality was affected by scan depth, tissue opacity, and light source characteristics. Longer wavelength (>1000 nm) light source OCT has been developed, which has several advantages, such as less scattering by opacity tissue and deeper penetration.^47,48 So for further research, applying OCT with a light source of longer wavelength will improve image quality and further extend the scan depth. Three-dimensional imaging technology of lens imaging for an overall view and digital modeling will be helpful to give a more comprehensive understanding of morphology and volume changing in lens regeneration. 

The lens regeneration pattern in our work is similar to previous studies in rabbit and rodent models.^{31,32,34,49–51} The difference between most of the other studies and our study is that we performed the capsulorrhexis at 3 mm, not minimally or sealed. The previous study shows that an intact lens capsular bag will enhance the shape, structure, transparency, and growth of the regenerated lens.³⁴ Preventing adhesions and wrinkles may improve the shape of the regenerated lens and in areas where the anterior capsule is missing, normal lens regeneration does not occur.^52–54 However, in our study, though the anterior capsule was impaired, lens regeneration with a fairly regular shape and translucency still could occur in a considerable amount of animals (Fig. 7), and some wrinkles and adhesions were reversible. Lois et al.,²⁶ who developed a PCO model by surgical technique with capsulorrhexis, also mentioned lens regeneration in all the animals in their study, but the observation period was not long enough to get a fully regenerated lens. Huang and Xie³¹ also used the same surgical technique²⁶ and found lens regeneration occurred. This implies that there may be a mechanism including lens capsular bag remodeling to help create a relatively intact or sealed environment, which will lead to better lens regeneration. While the anterior capsule was largely impaired (e.g., at initial stage of regeneration), the area of the CCC region was reduced after the posterior capsule folded; that might reduce the area for lens material from the peripheral area to confluence quickly. At month 2, the CCC region size of anterior capsule was much smaller than the initial size (Fig. 4K). This might be caused by the lens capsular bag shrinking or a newly synthesized capsule.³⁴ This, in accordance with a minimal or sealed incision, will lead to better lens regeneration. However, with the anterior capsule damaged, lens regeneration was not perfect, scars or more turbid material could be seen near the CCC region. Further research is needed to explore the mechanism that leads to better lens regeneration after lens capsule or capsular bag damage. Furthermore, stem cells are known to be crucial for tissue regeneration or repair. In our study, the LECs of central anterior capsule were removed during CCC procedures and lens regeneration still occurred. We speculated that new lens would arise from the residual LECs in the peripheral or equatorial region, where the niche of stem cells may be located.^55,56 The mechanism of stem cells and the lens capsular bag in the lens regeneration processes is still unknown, and needs further research. 

Lens regeneration research will help to explore novel therapy of cataract and PCO. On one hand, enhancing lens regeneration will recover lens structures and retain its transparency to produce a more natural lens that would have the functions of refraction and accommodation. On the other hand, a better lens regeneration might help to eliminate PCO forming after cataract surgery. 

In conclusion, in vivo UL-OCT imaging provides longitudinal data on lens regeneration process of the same subject, which is a useful tool and may allow one to follow and compare the lens regenerative process under different interventions or therapy after ECLE in rats. 

Acknowledgments

The authors thank Fan Lu, Xiang-Tian Zhou, Mei-Xiao Shen, Le-Le Cui, Zhe Xu, and Cong Ye for providing help in this study. 

Supported by the Natural Science Foundation of Zhejiang Province, China (Grant No. Y2110784) and the National Natural Science Foundation of China (Grant No. 81570869). 

Disclosure: K.-J. Zhou, None; Y.-N. Li, None; F.-R. Huang, None; Q.-M. Wang, None; A.-Y. Yu, None