Natural Course of Elschnig Pearl Formation and Disappearance

Abstract

Purpose.: To observe and analyze the daily and weekly changes in morphology and size of Elschnig pearls in eyes with after-cataract.

Methods.: Pseudophakic eyes with pronounced regeneratory posterior capsule opacification (PCO) were included in this prospective study. High-resolution retroillumination images were taken, and the changes in size and shape of Elschnig pearls were analyzed at baseline and at 2, 7, and 14 days later using dedicated imaging software.

Results.: In total, 6309 Elschnig pearls in 85 eyes of 77 patients were analyzed. On average, four pearls were found per square millimeter (range, 0.2–9.7/mm²; CV, 49.5%) with a mean cumulative area of 0.66 mm² (range, 0.03–2.18 mm²; CV, 66.7%) of all marked pearls per eye and follow-up, which was 3.8% of the analyzed area. The mean pearl size was 9630 μm² (range, 2390–33,745 μm²; CV, 61.0%) at baseline, which corresponded to a diameter of 174 μm. The mean change of a pearl per day was 583 μm² (range, 175-1631 μm²; CV, 55.4%) or a 6% change in area. In total, 36% (CV, 5.5%) of all pearls increased in size, 7% (CV, 8.6%) did not change, and 35% (CV, 6.7%) decreased in size during 1 week. Furthermore, 11% (CV, 6,3%) of all pearls newly appeared, and the same number disappeared (CV, 5.7%) within 1 week.

Conclusions.: Elschnig pearls disappear and appear within days. The degree of progression and regression varies greatly between eyes. Knowledge about Elschnig pearl turnover may be of importance for attempts to modulate lens epithelial regeneration or lens regrowth and for lens-refilling procedures.

Clinically, after-cataract or posterior capsule opacification (PCO) has two different components, one regeneratory and one fibrotic. Regeneratory PCO is more common and is the main reason for a decrease in visual function after implantation of an intraocular lens (IOL). ^1,2 It is thought to be caused by residual equatorial lens epithelium cells (LECs) that migrate and proliferate between the posterior capsule and the IOL, first forming syncytial PCO and later Elschnig pearls, leading to a decrease in visual acuity and loss of contrast sensitivity. ^3,4 Elschnig pearls, the main feature of regeneratory PCO, were thought to develop slowly and gradually over longer periods of time (months or years). ⁵ However, as we showed recently, ^6–8 Elschnig pearls undergo rapid turnover with progression and regression, within days or weeks. 

A better understanding of the natural course of PCO development may be of critical importance to tackle PCO, which is still the main obstacle for lens refilling or phaco-ersatz, the ultimate attempt to treat presbyopia. ⁹ Because it has not yet been possible to remove all LECs during surgery and complete depletion may compromise lens capsule integrity in the long-term, modulating LECs by pharmaceutical means is a possible solution. However, to do this, a better understanding of the dynamics of PCO change and Elschnig pearl turnover are necessary. 

The aim of this study was to observe the dynamics of change in the morphology and size of Elschnig pearls over a time span of weeks in eyes with PCO using dedicated software that allows exact measurement of each pearl over time. Additionally, we aimed to assess the variability of change between eyes in a cohort of 85 eyes to better describe the natural course of PCO development. 

Materials and Methods

Study Design

Patients were included from a continuous cohort referred for neodymium:YAG (Nd:YAG) laser capsulotomy treatment at our outpatient clinic at the Department of Ophthalmology of the Medical University of Vienna (Vienna, Austria). The inclusion criterion was the necessity of Nd:YAG laser for the treatment of regeneratory PCO (pearl-type) in one or both eyes. Exclusion criteria were tear film problems, corneal abnormality, and dense vitreous opacities to avoid artifacts in the retroillumination images. 

In each case the patient was informed about the option of immediate laser treatment. In every included case, the patient accepted the extension of the Nd:YAG laser treatment. At every follow-up the patient was asked whether further observation and, therefore, extension of the Nd:YAG laser treatment would be acceptable or whether the Nd:YAG laser treatment should be performed straight away. The study design was approved by the local ethics committee in light of these circumstances. All research and measurements followed the tenets of the Declaration of Helsinki. 

Image Acquisition and Analysis

High image quality was essential for this study because small structures must be clearly visualized to identify minor changes in PCO morphology that can be expected within the short observation period. To document PCO in the desired quality, we used a high-resolution digital retroillumination system that was specifically used for PCO assessment at our institution (digital coaxial retroillumination photography). The optical system consists of a slit-lamp (30 SL-M;) and a retrolux illumination module with illumination provided by an anterior segment flash pack through a fiberoptic cable (all from Carl Zeiss Meditec GmbH, Jena, Germany). This system allows coaxial illumination and high-contrast imaging. 

A digital color camera (NC2000; Kodak, Rochester, NY) was connected to the system. The images were directly transferred to a personal computer and imported into a graphics editing program (Photoshop 7.0; Adobe, San Jose, CA). The best images were saved to hard disc as tagged image files (TIF). 

Our image acquisition system produced high-quality digital images of the posterior capsule and was shown to have high reproducibility. ¹⁰ It is, therefore, a valuable method for PCO-related studies and for observing small changes during PCO development in vivo. A drawback of this method is the existence of Purkinje reflections in the images. ¹¹ However, these artifacts were negligible for this study because they affected only a small percentage of the area of the entire image. 

Digital retroillumination images were taken in a standardized fashion at baseline and at 2, 7, and 14 days later. On each occasion, the patients received topical phenylephrine 2.5% and tropicamide 0.5% for pupil dilation at least 30 minutes before examination. Then digital retroillumination images were taken with our image acquisition system in a standardized fashion: the patient was asked to look straight into the collimated light. Several images of the posterior capsule were taken (magnification, 12×), and the best images were stored on hard disc. 

After the acquisition of all images, the image of best quality (focus, exposure, position of reflexes) of each follow-up was selected and imported into the graphics editing program (Photoshop 7.0; Adobe). The program was used to adjust contrast and brightness of the follow-up images to a level similar to that of the baseline image. Then the images were converted from TIF format to portable network graphics format and imported into a numerical computing program (MATLAB 7.0.4; The MathWorks Inc., Natick, MA). With this numerical computing software and a dedicated plug-in (PearlTracer; Technical University of Graz, Austria; Fig. 1), sequential image series of each eye were created. Within these series single Elschnig pearls were traced from image to image. To mark the outlines of a pearl, the observer was able to choose among three features. In snake selection, the observer marked only a few points on the outline of the pearl, and the program detected the pearl border automatically. In diameter circle selection, the observer could choose a circle function and adapt it in size when the pearl was circular. In polygon selection, the observer marked the outline of the pearl manually. Analysis was always started from the first image to the last image of the series. Each pearl was identified by a number. Then the nonmarked pearls in the last image were traced backward toward the earlier images. Finally, the middle images were checked for nonmarked pearls, which would primarily have been newly appeared pearls. Pearls that were not visible in one or more images because of Purkinje reflexes were excluded from the analysis in all images. After marking all pearls inside the capsulorrhexis area, the program generated a spreadsheet file (Excel; Microsoft, Redmond, WA). The file contained the consecutive identification numbers of the marked pearls and their sizes in pixels on each follow-up image. To convert the values from pixels to square micrometer, three images of an implanted IOL with a known diameter were used as a reference. For statistical evaluation, the data were imported into analytical software (SPSS for Windows 12.0; SPSS Inc., Chicago, IL). 

Figure 1.

 

Screenshot of the custom-made software PearlTracer shows the same three magnified areas from three different retroillumination images of the same eye taken at different time points during follow-up to allow direct comparison and tracing of pearls over time. The small image depicts the entire nonmagnified image at one time point. Pearls are identified with a unique number and outlined. Numbers in the grid were used to detect the pearl position on the posterior capsule.

The pearls in the baseline image were grouped into three groups evenly divided by size, as follows: small pearls, 50 μm² to 3100 μm²; medium pearls, 3101 μm² to 7860 μm²; large pearls, 7861 μm² and greater. 

Mean pearl size and changes in all pearls sorted by size—such as increase, stagnation (changes less than ±5%), decrease, appearing, and disappearing—were analyzed. The cumulative area of pearls per eye and the correlation between the cumulative area per eye and the differences between the follow-ups and the number of pearls that appeared and disappeared in less than 2 weeks was analyzed. Furthermore, the mean pearl change per day per eye in square micrometers was calculated. To assess pearl density, the number of pearls in every eye was divided by the area of the entire marked area per eye and given in pearls per square millimeter. To calculate the average lifespan of pearls, the percentage of disappearing and appearing pearls per eye was analyzed. 

Finally, to assess the shape of Elschnig pearls, we analyzed the relationship between the area and the perimeter of each pearl and created two variables, solidity and roundness. The more frayed a pearl was, the lower was the solidity. Solidity should not be misunderstood as describing the mechanical properties of pearls but as a solely shape-describing variable. The more circlelike a pearl was, the higher was the roundness. 

Results

Patients who attended our clinic had undergone cataract surgery at least 1 year before Nd:YAG laser treatment. Photographs from 85 pseudophakic eyes with pronounced regeneratory PCO (pearl type) of 77 patients from three continuous studies were included for this study. Mean age of the patients was 72 years (range, 28–93 years). In all (85) cases, retroillumination images were taken on day 0, in 65 cases on day 2, in 59 cases on day 7, and in 77 cases on day 14. 

Altogether, 6309 pearls were marked, ranging from 21 to 151 pearls per eye. An example is shown in Figure 2. On average four pearls were found per square millimeter (range, 0.2–9.7/mm²; coefficient of variation [CV], 49.5%). The 25% of eyes (lower quartile) with the lowest pearl density showed a higher number of appearing pearls (12 appearing pearls per eye [range, 0–29 pearls]) and a lower number of disappearing pearls (9 disappearing pearls per eye [range, 2–16 pearls]) during 2 weeks compared with eyes with a high pearl density (4 appearing pearls per eye [range, 0–12 pearls] and 13 disappearing pearls per eye [range, 2–40 pearls]; P = 0.0001). 

Figure 2.

 

Retroillumination images were taken at baseline (left), 1 week later (middle), and 2 weeks later (right). Examples of an increase in pearl size (white arrow) and an increase followed by a decrease in pearl size (black arrow) are depicted in the same eye.

The mean cumulative area of all marked pearls per eye and follow-up was 0.66 mm² (range, 0.03–2.18 mm²; CV, 66.7%). Over the follow-up period, the mean cumulative area of all marked pearls changed from 0.66 mm² at day 0 (range, 0.03–2.20 mm²; CV, 66.7%) to 0.60 mm² (range, 0.03–1.88 mm²; CV, 073.3%), 0.68 mm² (range, 0.06–1.95 mm²; CV, 100%), and 0.62 mm² (range, 0.09–1.82 mm²; CV, 66.1%) at days 2, 7, and 14, respectively (Fig. 3, top). 

Figure 3.

 

Change (percentage) in mean cumulative pearl area per eye (top) and change in mean pearl size (μm²) per eye (bottom) from baseline to day 2 (left) and to day 14 (right).

Concerning pearl size change within 1 week, 36.3% (CV, 5.5%) of all pearls increased in size, 7.0% (CV, 8.6%) did not change in size, and 34.9% (CV, 6.7%) decreased in size. Furthermore, 10.5% (CV, 6.3%) of all pearls had newly appeared, and 11.3% (CV, 5.7%) of all pearls had disappeared during 1 week (Fig. 4). On average, 27 pearls per eye (range, 0–76 pearls; CV, 8.9%) decreased, 6 pearls (range, 0–24 pearls, CV, 10.0%) did not change in size, and 29 pearls per eye (range, 0–68 pearls, CV, 6.9%) increased in size during 1 week. Furthermore, 7 pearls (range, 1–23 pearls, CV, 10.0%) per eye appeared, and 7 pearls (range, 0–20 pearls, CV, 8.6%) per eye disappeared during 1 week. Table 1 summarizes the changes in all pearls and the dependence on pearl size during 1 week. 

Figure 4.

 

Percentage of appearing and disappearing pearls per eye within 1 week.

Table 1.

 

Morphologic Changes in All Pearls within One Week

Pearl Size	Decrease (%)	Stagnation (±5%)	Increase (%)	Disappearing (%)	Appearing (%)	Total (%)
Small	183 (14.4)	65 (5.1)	489 (38.6)	131 (10.3)	400 (31.5)	1268 (100)
Medium	517 (40.8)	104 (8.2)	499 (39.4)	148 (11.7)	0 (0)	1268 (100)
Large	627 (49.4)	97 (7.6)	393 (31.0)	151 (11.9)	0 (0)	1268 (100)
Total	1327 (34.9)	266 (7.0)	1381 (36.3)	430 (11.3)	400 (10.5)	3804 (100)

Mean pearl size changed from 9630 μm² at day 0 (range, 2390–33745 μm²; CV, 61.0%) to 10306 μm² (range, 2223–32568 μm²; CV, 61.3%), 10972 μm² (range, 1332–30717 μm²; CV, 59.3%), and 10069 μm² (range, 2871–33466 μm²; CV, 63.4%) on days 2, 7, and 14, respectively (Fig. 3, bottom). 

The solidity of pearls decreased slightly from 0.85 (0.40–0.98; CV, 11.8%) at baseline to 0.83 (0.41–0.98; CV, 13.3%) at the 1 week follow-up (Fig. 5). The roundness decrease was constant, with 0.06 (0.03–0.07; CV, 0.17%) at baseline and at the 1 week follow-up. The upper quartile of all eyes with the highest pearl density showed a higher solidity (mean, 0.8; range, 0.7–0.9) and a higher roundness (mean, 0.07; range, 0.06–0.07) than the eyes in the lower pearl density quartile (mean, 0.7; range, 0.5–0.8; mean, 0.06; range, 0.03–0.07, respectively). 

Figure 5.

 

Box plot of the solidity of all pearls. The higher the solidity, the less frayed and the more regular the pearl, as also illustrated by the sample photographs. Representative pearls chosen at each quartile are depicted for better understanding. The uppermost pearl has a high solidity of 1, whereas the lowermost pearl has a frayed shape.

At baseline and on days 3 and 7, mean pearl sizes of 2720 μm², 5300 μm², and 6150 μm², respectively, were found for newly appearing pearls. The solidity of 0.977, 0.980, and 0.976 and the roundness of 0.0727, 0.0733, and 0.0721 were higher than those of pearls that showed no or little change. At baseline and on days 3 and 7, mean pearl sizes of 8100 μm², 8500 μm², and 7400 μm², respectively, were found for disappearing pearls. The solidity of 0.974, 0.974, and 0.972 and the roundness of 0.0726, 0.0726, and 0.0719 were also higher than those of pearls that showed no or little change. 

Mean change in the size of a pearl per day was 583 μm² (175–1631μm²; CV, 55.4%). There was no correlation between the pearl density and the daily change in pearl size (r = 0.10; P = 0.00). However, daily change in pearl size correlated with solidity (r = 0.52; P = 0.00) and roundness (r = 0.49; P = 0.00; Fig. 6). 

Figure 6.

 

Scatterplot of daily change in pearl size (μm²) and solidity. A low solidity value indicates a frayed, more irregular shape of the pearls.

Both the appearance and the disappearance of pearls during the period of follow-up occurred in a linear fashion (Fig. 7). The probability that a pearl disappeared during a time period of 1 week was 11.3% (CV, 5.3%) and that a new pearl appeared was 10.5% (CV, 6.7%). Extrapolating from this linear relationship, all pearls that were present at baseline would have disappeared after 8.9 weeks, and it would have taken 9.5 weeks for a completely new set of pearls to be formed. The calculated mean lifespan of a pearl would be on the order of 19 weeks. Pearls with a lifespan of less than 2 weeks were observed in only 1% of all pearls. The largest size these short-lived pearls attained was 9255 μm² on average. All other observed pearls existed longer than 2 weeks. Taking into account the rapid morphologic changes found, we decided to focus on the size and shape changes of Elschnig pearls during the shorter time range of 1 week instead of 2 weeks. 

Figure 7.

 

Percentage of appearing and disappearing pearls over the period of follow-up shown for each eye (left) and for all pearls together (right).

To create a representative simulation of the size and shape characteristics of a pearl during its lifetime, mean values for size and shape were calculated, and representative pearls from the dataset were selected for illustration (Fig. 8). On the first day the pearl's appearance is small, solid, and round. During the next days the pearl increases in size but stays round and solid. Then the pearl's solidity and roundness decrease, and several morphologic changes are possible. In the schematic illustration (Fig. 8), a questionable separating, or dividing, pearl is shown that was thought to be seen in several instances in the dataset. In the weeks and days before disappearing, the pearl can undergo two different morphologic changes. Some pearls increase in size within this time period before disappearing, whereas other pearls decrease in their last days before they disappear. Whether they increase or decrease in size, usually their solidity and roundness increase again before they disappear. 

Figure 8.

 

Illustration of the typical morphologic changes of a pearl during its lifespan. Size and shape characteristics were taken from different pearls to fit the mean values found in the dataset. Right: two possible pathways at the end of a pearl's lifetime are depicted: growth to a large pearl and then disappearance (a) or shrinkage of the pearl until undetectable (b). In addition, the questionable fragmentation of a pearl is shown (c).

Neither IOL material nor age or sex (r ² < 0.4 in all cases) had any significant influence on the behavior of pearls, whether concerning change in size or morphology. 

Discussion

In this prospective study, morphologic short-term change of 6309 Elschnig pearls in 85 eyes were analyzed. We found pronounced changes of pearls between every follow-up examination in every eye. The degree of increase and decrease in pearl size and the number of disappearing and newly formed pearls was heterogeneous among eyes and among patients. On the one hand, there were eyes with a highly active progression and regression of pearls (i.e., a high turnover of pearls), and on the other hand there were eyes with little progression and regression. Even the shape of the pearls was variable among eyes, but, in general, pearls that were newly formed and pearls that were soon to disappear showed higher solidity and roundness than pearls during the middle part of the lifespan. 

Concerning size, newly appearing pearls were always small, but pearls that disappeared ranged from large to small, indicating a different mechanism for loss of pearl volume than “slow death.” Osmotic variations seem likely to be the mechanism, with rapid change in pearl size instead of pearl death. However, an extreme decrease in pearl volume does not necessarily mean the complete disappearance of the pearl because it may still be present in a very small size that is not identifiable in retroillumination photography. 

We found similar changes in our previous studies in which small areas of pearls were assessed semiquantitatively by a masked examiner. ^6–8 In the first study we subjectively assessed the overall change of pearl size and pearl number in 19 pseudophakic eyes during time intervals of 2 weeks using a grid system overlaid on the images. In approximately 70% of the eyes, pearls increased in size during 2 weeks. A reduction of pearl size or even a total disappearance of pearls was also found to a similar extent, as in this analysis. In the second study, ⁶ we investigated 27 pseudophakic eyes and reduced the time period between the follow-up examinations to single days and again subdivided the images into small standardized areas for subjective analysis by overlying a fine grid onto the images. Even in these short periods of time, pronounced changes in pearls were found, showing size increases or decreases and even the appearance or disappearance of pearls in numerous cases. The third study ¹² was a prospective, randomized, double-blinded, placebo-controlled, two-way-crossover study to assess the effect of a combination of topical steroidal and non-steroidal anti-inflammatory drops on the formation of pearls in 39 eyes. To quantify the increases and decreases in pearl size we used the PearlTracer software, which was used for the first time in this study. Evaluation with the PearlTracer software showed high reproducibility. However, the anti-inflammatory treatment had no effect on the short-term changes of pearls. ⁸  

Single cases of spontaneous regression or disappearing Elschnig pearls were also reported by Nakashima et al. ¹³ and Caballero et al. ¹⁴ The latter authors assumed that the most likely cause of pearl regression was cell death by apoptosis. They hypothesized that after several years of proliferation and migration, cells may lose their capacity to proliferate and begin to die by apoptosis. This mechanism would explain the significant reduction in the number of posterior capsulotomies performed more than 4 years after cataract surgery. ¹⁵ Posterior capsulotomy seems to influence the formation of the remaining PCO structures, especially the formation of pearls. Cases of proliferation and ensuing regression of Elschnig pearls after Nd:YAG capsulotomy have been published. ^15–18  

To detect size changes of Elschnig pearls within the short time of days and weeks, high image quality was essential for this study. A sophisticated optical setup, and reproducible image acquisition ¹⁰ allowed us to examine the change of many pearls in detail. Dedicated software enabled us to analyze a large number of pearls in a reproducible way. However, although the process of analysis was semiautomated, the expenditure of time was high. For trials with a larger numbers of patients, the efficiency of image processing would have to be improved. 

Several questions remain to be answered. First, what are the factors influencing the dynamics of change of PCO? We found great variability between patients, as seen in previous studies. ^6,7 The factors IOL material, age, and sex did not have an effect on pearl turnover in this study. However, the number of eyes included was possibly too small to detect these influencing factors. Second, we have found that the pearls that disappear are typically round and have a smooth shape. Similar characteristics are found with new and “growing” pearls. This raises the question whether these pearls are different subsets from those that are more stationary, show little change, and appear to have a more frayed shape that also deviates from the round form. Or are this change and transition in shape actually part of the lifecycle of a pearl as we have indicated in Figure 8. Third, does the actual fragmentation or division of a large and often nonround, frayed pearl into two distinct pearls happen? This could imply that mitosis takes place if these were large cells. 

The question remains, what is an Elschnig pearl morphologically? Cowan and Fry ¹⁹ published their histologic observation describing Elschnig pearls as giant cells formed by a regenerative attempt of lens epithelial cells. Fagerholm ²⁰ compared lens fiber cells with Elschnig pearls. His analysis showed that Elschnig pearls may be cells that are differentiating to lens fiber cells. However, contrary to lens fiber cells, Elschnig pearls still contained a nucleus and a few cell organelles. These so-called lens-fiber like cells may form extracellular vacuoles that appear as transparent spheres. These vacuoles were found to be surrounded by very thin lens fiberlike cells. This theory was based on a model by Sakuragawa ²¹ that the intact lens may respond, when exposed to extracellular fluid, by increasing the transport of sodium, and thereby water, into the extracellular space with the resultant formation of a cystlike cavity. 

Kappelhof et al. ^22–24 described the Elschnig pearls in rabbits as subcapsular epithelial cells originally from the Soemmering's ring that escape and migrate toward the center of the posterior capsule. Transmission and scanning electron microscopy showed two different kinds of cells between the IOL and the posterior capsule: One type of cell contained an indented nucleus with numerous patches of chromatin, large cisterns of endoplasmic reticulum, numerous clear vacuoles, and a normal set of cell organelles. The second type of cell consisted of a round-to-oval nucleus, unfolded chromatin and a clear-cut nucleolus, few cytoplasmic organelles, and a homogeneous cytoplasmic matrix. 

A contrasting theory by Jongebloed et al. ²⁵ is based on transmission electron microscopy after collecting Elschnig pearls with a glass cannula onto a Millipore filter. In his study, no cell membrane remnants were found, indicating that Elschnig pearls are biophysical products of a slow process of degradation of lens fibers and having no cellular origin. One argument that confirms this theory is that the TEM images showed erythrocytes close to Elschnig pearls. These erythrocytes were deformed because of the osmium staining used before TEM, whereas Elschnig pearls retained their shape. The explanation was that Elschnig pearls are products of the ballooning of the cell membrane, emerging from the cytoplasm of the lens fibers. Similar formations are found in cataractous lenses. The origin of the Elschnig pearls could be the cell membrane of the degenerating lens fiber that consists of unsaturated lipids. This would explain why the osmotic staining had no effect on the pearl size and shape; the real cells were deformed. 

Marcantonio et al. ²⁶ describe the phenomenon of epithelial mesenchymal transition of LECs as the cellular substrate of PCO including Elschnig pearls driven by cytokines, such as TGF-β. 

Therefore, the literature to date has conflicting theories regarding what Elschnig pearls actually are. Based on our clinical findings, there are some characteristics that speak in favor of the pearl being a cystlike structure. The first is the rapid change in size seen in some cases and especially the disappearance of large pearls within few days. Second, the very large size of some of the pearls would make these extremely large cells. The third characteristic is the very low rate of fragmentation/division of Elschnig pearls. In addition, in these cases of questionable fragmentation/division, it is feasible that a large pearl has decreased in size and that a small neighboring pearl undergoes a concomitant increase in size. 

On slit-lamp examination using indirect illumination and a method to assess the refractive index of round structures, as described by Brown, ²⁷ Elschnig pearls appear to be filled with material of higher refractive index than the surrounding tissue. This would speak in favor of a solid and cellular origin of pearls or, alternatively, a cystlike formation filled with lens fiber material. 

To influence the dynamic behavior of PCO development is of great relevance because PCO is still a major restricting factor in lens-refilling procedures in animals to allow restoration of accommodation. ⁹ In addition, the difficulty of Nd:YAG laser treatment in developing countries underscores the importance of further investigating alternative methods of treating PCO. 

In conclusion, this study showed that Elschnig pearls disappeared and appeared within days and underwent significant changes in size and shape in short periods of time. The degree of progression and regression varies greatly among patients. 

Footnotes

 Disclosure: O. Findl, None; T. Neumayer, None; N. Hirnschall, None; W. Buehl, None

The authors thank Stephen Tuft, FRCO, for the fruitful discussion on the clinical appearance of pearls.