Introduction

Nowadays cataract removal with implantation of intraocular lens (IOL) in the capsular bag is the most frequent ophthalmic surgical procedure performed. The visual benefit obtained after this intervention may progressively suffer a decrease due to posterior capsular opacification (PCO). This long-term complication is caused by proliferation and centripetal migration (on the posterior capsule, as far as the blockage of visual axis) of lens remnant epithelial cells left in both the lens equator and anterior capsular bag following cataract extraction in the majority of cases.1

Significant differences exist among the published rates of PCO, yet a meta-analysis published in 1998 reported that overall, 25% of patients undergoing extracapsular cataract extraction or phacoemulsification developed visually significant PCO within 5 years of the operation.2 The defect of visual acuity caused by PCO can usually be remedied by laser surgery with neodymium : YAG (Nd : YAG) capsulotomy to create an opening in the posterior lens capsule.

In spite of recent advances in imaging diagnostic methods,3, 4, 5 visual field analysis through white-on-white automated perimetry is a widely used technique that continues as a standard exploration to detect and follow-up glaucoma6 and neuro-ophthalmological diseases.7, 8 Opacities in the anterior eyeball pole, such as cataracts or PCO, could be a confusing factor that may lead to an incorrect interpretation of automated perimetry, even when it is not uncommon to encounter patients who are affected, or are suspected to be affected, by various entities: PCO and glaucoma, or PCO and optic neuropathy for instance. The clinician must decide how much visual damage is caused by PCO, and how much is caused by the other disease.

Clinically, the cataract and PCO aspect upon slit-lamp examination is completely different: while the majority of cataracts present a homogeneous opacity, most common PCOs appear to have an unpredictable conformation, so a dissimilar perimetric behaviour is suspected.

The influence that cataracts have on automated perimetry has been well studied.9, 10, 11 Concerning PCO however, some studies have been undertaken by analysing its effects on visual acuity, contrast sensitivity, and glare, as well as on the improvement of these parameters after Nd : YAG capsulotomy,12, 13, 14, 15, 16 and also the influence of pseudophakia without PCO on kinetic17, 18, 19, 20 or static perimetry.21, 22 Nonetheless, no studies have analysed the perimetric behaviour of PCO.

In this study, we compared the characteristics of visual fields obtained with white-on-white automated perimetry from patients affected by PCO before and after Nd : YAG capsulotomy, and we determined the effect of PCO on the visual field.

Materials and methods

We contemplated all pseudophakic patients who came to our office complaining of decrease of visual acuity, glare, or other types of progressive visual discomfort for their possible inclusion in this study. Only those patients who underwent previous uneventful cataract surgery with acrylic IOL in the capsular bag were recruited, where the cause of visual alteration was exclusively due to PCO (patients with other pathology of the cornea, vitreous, macula, or optic nerve were excluded) after a complete ophthalmologic assessment, which included the following in this order: BCVA, slit-lamp examination, applanation tonometry, and retinal exam under pupilar dilation (tropicamide 1% and phenylephrine 10%). If PCO was detected after slit-lamp examination, the patient waited at least 15 min, and subsequently, a visual field examination through automated perimetry white-on-white SITA standard threshold strategy with Humphrey Field Analyzer II (model 740, Humphrey Instruments, Inc., Dublin, CA, USA) was performed by the same experienced perimetrist on both eyes, the fellow eye followed by the PCO-affected eye. Goldmann size III target was used. The results of this first visual field test were not considered for data analysis. The remaining examinations (applanation tonometry and funduscopy) were performed after automated perimetry.

Selected patients returned 1 week later, and they underwent identical visual field examination. Only those patients who did not surpass 20% false-positive or false-negative responses on visual field examination remained in the study. Forthwith, patients underwent mydriasis (tropicamide 1% and phenylephrine 10%) and Nd : YAG capsulotomy by the same surgeon. The laser treatment was initiated off axis in a vertical line across the centre, followed by a line on the horizontal axis to form a cross. The minimum amount of energy was used. In general terms, the aim was to create a capsulotomy size that approximately measured the minimum size required to exceed the pupillary diameter under scotopic conditions. Surgical eyes received one drop of apraclonidine 1% immediately after Nd : YAG, and 0.2% brimonidine twice daily for 7 days.

During the next visit, which was scheduled between 1 and 8 weeks after capsulotomy, an identical complete ophthalmic examination was performed including a control automated perimetry (with the same characteristics and the same exclusion criteria as the previous test).

At the end of this prospective selection process, 30 eyes of 30 patients were affected by PCO without any other significant ocular pathology, carrying acrylic IOL in the capsular bag. Four of them were excluded in the first taken into account visual field examination (none in the second) as they exceeded 20% of false-positives or false-negatives. Therefore, 26 eyes of 26 patients accomplished all the requirements described above. An informed consent was obtained from each patient before enrolment. All procedures conformed to the tenets of the Declaration of Helsinki. The study was approved by the local Ethics Committee.

Results

Statistical analysis was performed between preoperative and postoperative parameters by means of the paired-samples Student's t-test.

Out of these 26 patients, 12 were men and 14 were women. The mean age was 67.65±8.03 years (range 51–83 years). The mean period between cataract extraction and development of visually disturbing PCO was 29.43±8.17 months (range between 13 and 43 months). The 26 PCO-affected eyes were 13 right eyes and 13 left eyes.

The BCVA before and after Nd : YAG capsulotomy were 0.35±0.11 and 0.84±0.14, respectively, where a mean improvement of 4.92 lines on the Snellen chart after capsulotomy was observed (P<0.001, paired-samples Student's t-test) (Figure 1).

Figure 1
figure 1

Summary of BCVA of PCO-affected eyes (treated with Nd:YAG laser) before and after capsulotomy.

Full size image

No significant difference (P=0.83, paired-samples Student's t-test) between mean preoperative (17.5±2.4) and postoperative IOP (17.42±2.31) was observed.

Postoperatively, not only the MD, but also the pattern standard deviation (PSD) revealed a statistically significant improvement after capsulotomy in 3.46 and 1.38 decibels (dB), respectively. However, the same indices of fellow eyes remain unchanged (Table 1).

Table 1 Preoperative and postoperative global indices (n=26)
Full size table

Detailed results of PCO-affected eyes (treated) (Figures 2 and 3) and fellow eyes (untreated) (Figures 4 and 5) are represented in the charts.

Figure 2
figure 2

Summary of MD values of PCO-affected eyes (treated with Nd : YAG laser) before and after capsulotomy.

Full size image
Figure 3
figure 3

Summary of PSD values of PCO-affected eyes (treated with Nd : YAG laser) before and after capsulotomy.

Full size image
Figure 4
figure 4

Summary of MD values of fellow eyes (control) before and after capsulotomy.

Full size image
Figure 5
figure 5

Summary of PSD values of fellow eyes (control) before and after capsulotomy.

Full size image

No retinal complications (such as peripheral degenerations, retinal tears or retinal detachments) were found either before nor after Nd : YAG was applied.

Bivariate associations analyses between BCVA, MD and PSD values (precapsulotomy and postcapsulotomy) were assessed using the Spearman correlation coefficient. Any differences showing a P-value of less than 0.05 were considered as statistically significant. There was a significant association among all variables to a greater or lesser extent (Table 2).

Table 2 Spearman correlations between BCVA, MD and PSD before and after capsulotomy
Full size table

Simple linear regression analyses were also performed to assess the predictive capability of the different parameters before the capsulotomy. An increase of the BCVA precapsulotomy entailed an improvement of both the MD postcapsulotomy, and also the PSD postcapsulotomy to a lesser extent (Figures 6 and 7). Likewise, the worst MD and PSD values before the Nd : YAG application predicted the worst BCVA values after the capsulotomy (Figures 8 and 9).

Figure 6
figure 6

Scatter plot of BCVA precapsulotomy (independent variable) MD postcapsulotomy measurements (dependent variables). Linear regression function is represented. R2 indicates the proportion of variation in the dependent variable (represented on the y-axis) that can be explained by the independent variable (represented on the x-axis).

Full size image
Figure 7
figure 7

Scatter plot of BCVA precapsulotomy (independent variable) vs PSD postcapsulotomy measurements (dependent variables). Linear regression function is represented. R2 indicates the proportion of variation in the dependent variable (represented on the y-axis) that can be explained by the independent variable (represented on the x-axis).

Full size image
Figure 8
figure 8

Scatter plot of MD precapsulotomy (independent variables) versus BCVA postcapsulotomy (dependent variable). Linear regression function is represented. R2 indicates the proportion of variation in the dependent variable (represented on the y-axis) that can be explained by the independent variable (represented on the x-axis).

Full size image
Figure 9
figure 9

Scatter plot of PSD precapsulotomy (independent variables) versus BCVA postcapsulotomy (dependent variable). Linear regression function is represented. R2 indicates the proportion of variation in the dependent variable (represented on the y-axis) that can be explained by the independent variable (represented on the x-axis).

Full size image

Discussion

PCO shares many characteristics with cataracts as both interfere with light transmission to the retina. Visual acuity improves once this blockage effect has been suppressed (almost 5 lines after capsulotomy in this study).

As far as IOP is concerned, no change was indicated in our study including a follow-up of between 1 and 8 weeks after capsulotomy. These results are in accordance with those obtained by Keates et al, since they found that the majority of IOP returned to the pretreatment level in 1 week post capsulotomy.23

In relation to visual field examination, we used the programme 24-2, which is accepted nowadays as the standard, given that it contains 99% of the information provided by 30-2 but is 25% faster, so it involves less fatigue, and therefore less short-term fluctuation.24, 25 The visual fields inclusion criterion was established in less than 20% for both false-positive and false-negative errors. This is more demanding than the perimeter manufacturers recommendations (Humphrey Instruments, Inc., Dublin, CA, USA), as they suggest less than 33% (for both types of errors) in order to obtain a reliable examination.

Otherwise, fixation losses were not considered as an inclusion criterion since there are circumstances in which a high rate of fixation losses could appear when the patient has actually maintained a good fixation.26

Results from the first automated perimetry were not taken into account as the aim of this test was to eliminate the learning effect. Although most normal patients will produce reliable results even in their first perimetric test, a significant minority of subjects will improve considerably with practice.27 This learning effect occurs mainly between the first and second visual fields.28

PCO-affected eye perimetry was performed after fellow eye perimetry in all patients in order to standardize the learning effect transfer from the first to the second examined eye within the test.

A visual field of the fellow eye (control eye) was also performed in each patient with the purpose of quantifying the possible learning effect between the considered pre- and post-laser automated perimetry. No significant learning effect was observed in our study (MD increase of 0.09 dB, P=0.387; PSD decrease of 0.11, P=0.127) in the control eyes. We could, therefore, conclude that any MD and PSD changes in the PCO-affected eyes (hypothetically influenced by a similar learning effect) is due to capsulotomy.

Although visual acuity, contrast sensitivity, and glare have been demonstrated to be satisfactory in capsulotomies as small as 1.5 mm in diameter, even slightly decentred ones,29 in our study the size of capsulotomy was established as being larger than the pupillary diameter in dark conditions30 since a smaller capsulotomy size could affect the results of automated perimetry.31

The test–retest interval (maximum eight weeks) was short enough to postulate that no true, detectable visual field progression occurred.

MD is the average measure of how depressed the patient's visual field is (compared with a normal person of the same age). Several researchers reported that MD improved after cataract surgery (3.92 dB,9 1.68 dB,10 3.52 dB11). Similarly, there was amelioration of MD after capsulotomy, according to our results (3.46 dB, P<0.001).

PSD is a measure of how different the adjacent points are on a visual field. If an area is focally depressed, the PSD will rise given the great difference that exists between the points in the scotoma and their normal adjacent points. PSD remains unchanged after cataract removal (it worsens by 0.54 dB (P=0.09)10 or by 0.2 (P=0.24),9 and it improves by 0.04 (P=0.93)11). In relation to PCO however, PSD improves by 1.38 dB after capsulotomy in our study (P<0.001). This modification could be explained by the special PCO properties concerning reflection, refraction, and diffraction. PCO optical translucent imperfections presumably may cause irregular light-scatter within the eye, resulting in a combination of differently under-illuminated retinal areas and in an increase of PSD. Yet when these irregularities have been eliminated through capsulotomy, retinal illumination would be more uniform, so PSD decreases.

As the clinical slit-lamp examination impression suggests, and as the analysis of automated perimetry confirms, cataracts depress an automated visual field fairly uniformly, so they constitute homogeneous opacities. However, PCOs depress the visual field heterogeneously. They are, therefore, demonstrated as being polymorphous opacities that may even simulate pathological patterns, such as glaucoma arcuate scotoma, which are susceptible to elimination after capsulotomy (Figures 10 and 11).

Figure 10
figure 10

Visual field of right eye (PCO-affected) before and after capsulotomy in patient 11. Prelaser perimetry simulates an arcuate glaucomatous defect that disappears after laser therapy. Note important changes in DM and DSM. DM=MD, DSM=PSD, PHG=Glaucoma Hemifield Test, FL=Fixation loss, FN=False-negative, FP=False-positive.

Full size image
Figure 11
figure 11

Visual field before and after capsulotomy of fellow eye (left eye, where no capsulotomy was performed) in patient 11. Note there is no significant change in DM or DSM. This eye is affected by a mild corticonuclear cataract. DM=MD, DSM=PSD, PHG=Glaucoma Hemifield Test, FL=Fixation loss, FN=False-negative, FP=False-positive.

Full size image

As a result, it is necessary to consider the presence of PCO in the interpretation of visual fields in all pseudophakic patients since this entity may mislead the results, or may even mimic the perimetric behaviour of other pathologies.

The correlation results revealed a significant association among BCVA, MD and PSD, both prior to and after capsulotomy. The strong correlation between the BCVA and PSD precapsulotomies stands out. As is known, pear-type PCO produces a greater visual acuity decrease than the fibrosis-type PCO.1 Perhaps the PCOs with the worst visual acuity in this study are those with a greater rate of pearl formation, optical irregularities that would lead to an increase of PSD. Further study is necessary in this particular sense.

Linear correlation analyses indicate that BCVA precapsulotomy may better predict MD postcapsulotomy than PSD postcapsulotomy. Alternatively, MD precapsulotomy and PSD precapsulotomy are predictor parameters of BCVA postcapsulotomy.

All these conclusions are related to healthy subjects. However, more studies are necessary to discover whether these conditions are different in glaucomatous patients.