Does Cataract Extraction Significantly Affect Intraocular Pressure of Glaucomatous/Hypertensive Eyes? Meta-Analysis of Literature

Background and Objectives: This study aimed to evaluate the effect of cataract extraction on intraocular pressure at 6, 12, and 24 months and their difference compared to the baseline in diverse glaucoma subtypes. Materials and Methods: We carried out research in the MEDLINE, Cochrane Library and EMBASE databases, as of April 2022 for relevant papers, filtered according to established inclusion and exclusion criteria. The meta-analysis evaluated the Mean Reduction and relative Standard Error in these subpopulations at predetermined times. A total of 41 groups (2302 eyes) were included in the systematic review. Due to the significant heterogeneity, they were analysed through a Random Effects Model. Results: We obtained these differences from baseline: (1) Open Angle Glaucoma at 6, 12 and 24 months, respectively: −2.44 mmHg, −2.71 mmHg and −3.13 mmHg; (2) Angle Closure Glaucoma at 6, 12 and 24 months, respectively: −6.81 mmHg, −7.03 mmHg and −6.52 mmHg; (3) Pseudoexfoliation Glaucoma at 12 months: −5.30 mmHg; (4) Ocular Hypertension at 24 months: −2.27 mmHg. Conclusions: Despite a certain variability, the reduction in ocular pressure was statistically significant at 6, 12 and 24 months in both Open Angle Glaucoma and Angle Closure Glaucoma, the latter being superior. Data for Pseudoexfoliation Glaucoma and for Ocular Hypertension are available, respectively, only at 12 months and at 24 months, both being significant.


Introduction
Over 20 years ago, by common accord between the World Health Organization and the International Agency for Prevention of Blindness, the initiative "Vision 2020: The Right to Sight" was created, the results of which were recently published and reveal that, in adults aged 50 or over, the two main causes of blindness or moderate to severe reduction in visual acuity are, respectively, Cataract (45.4%) and Glaucoma (11%) [1].
These are two diseases which often coexist in a single patient and influence one another; the development and progression of a cataract can contribute to alterations in the drainage of aqueous humour while traditional filtering surgeries used to treat glaucoma can lead to the formation or worsening of the opacification in phakic patients [2].
Cataract extraction surgery is one of the best in terms of cost/benefit analysis and still is the most performed surgical procedure every year in several countries [3].
The objective of this meta-analysis is to demonstrate the correlation between cataract surgery and the evolution of the most crucial and the only modifiable factor in glaucoma, Intraocular Pressure (IOP).

Materials and Methods
A total of 34 studies, including 39 groups, from 1995 to 2019, were included in the meta-analysis (Table 1) based on the following inclusion and exclusion criteria.
Studies providing data on IOP pre-phacoemulsification and post-phacoemulsification.

2.
Studies approved by an institutional revision group or by an ethical committee.
Papers not available in English.

2.
Papers not available in a digital format.
Studies conducted on patients under 18 years of age. 5.
Preceding or concurrent trabeculectomy, other major ocular surgery or relevant illness.6.
A follow-up period of less than 12 months.7.
Relevant study arm with less than 15 eyes analysed.8.
Studies on MIGS without an arm treated only with phacoemulsification. 9.
Different subtypes of glaucoma included in the same arm.
The result of primary interest in our meta-analysis was the variation in IOP at times t1 (6 months), t2 (12 months) and t3 (24 months) compared to t0 (pre-surgical baseline).

Literature Research Method
We collected publications on the effects of cataract surgery in patients with glaucoma or ocular hypertension via a thorough research on the MEDLINE, Cochrane Library and EMBASE databases up to 1 April 2021.This systematic review was not registered in the international prospective register of systematic reviews (PROSPERO).In conjunction, we picked the search keywords and inclusion/exclusion criteria based on systematic reviews and previous meta-analyses.Our updated search includes relevant studies not included in previous review works.The keywords used in the research were ("Glaucoma, Open-Angle" OR "Glaucoma, Angle-Closure" OR glaucoma) AND ("Phacoemulsification" OR "phacoemulsification") AND ("Intraocular Pressure" OR "intraocular pressure") AND ("Ocular Hypertension" OR "ocular hypertension") AND ("Glaucoma, Pseudoexfoliative") AND ("Cataract Extraction" OR "cataract extraction") AND ("Cataract Surgery" OR "cataract surgery").
Results were initially selected depending on the relevance of the title and abstract and subsequently on the adherence to the inclusion/exclusion criteria.
The study arms were divided in Primary Open Angle Glaucoma (POAG), Angle-Closure Glaucoma (ACG), Pseudo-Exfoliative Glaucoma (PXG), and Ocular Hypertension (OH).Some authors included, in the same arm, POAG and OH, or POAG and PXG, or ACG and angle closure disease (ACD), where ACD does not imply the presence of glaucomatous damage to the optic nerve.As a result of the impossibility of separating the sub-groups, these were intentionally omitted from the studies selected for our analysis.
In cases where the studies reported and additional treatment arm, e.g., MIGS combined with cataract extraction surgery, only the phacoemulsification arm was included in this study.Certain studies reported different glaucoma subtypes within the same group and were thus excluded from the analysis.Other studies were excluded due to being extrapolated from the same dataset of previously published papers, thus risking that redundant data would be given to the statistical software to be analysed.
During the literature research, we also found that some studies that did not exclude eyes which underwent, prior to or during the follow-up period, to ocular surgeries or laser procedures (e.g., Selective Laser Trabeculoplasty) and were thus ruled out.
The result of this literature review, as represented in the flow diagram (Figure 1), is a total of 39 groups (20 POAG, 11 ACG, 4 PXG e 4 OH) that were selected from the 34 studies.In total, 11 of the selected arms were retrospective in nature, while 28 were prospective.Overall, 2227 eyes were analysed at the start of the follow-up.

Analysis and Data Synthesis
A meta-analysis is a quantitative technique allowing, from a clinical and statistical standpoint, for researchers to combine different studies on a single clinical issue to reach a single conclusive result with a more consistent statistical power.
The meta-analysis is completed through Statsdirect 3.2 (StaTsdirect, Wirral, UK).The data we focus on are average IOP pre and post surgery and their standard deviation.In the studies where the confidence interval is not reported, the standard error and/or p-values are extracted to calculate average and standard error of the difference compared to baseline (t0).
Using a random effects model, we can attempt to estimate the average of the true Some studies included a wash-out period as part of their methodology, i.e., the IOP levels were measured after a temporary suspension period of glaucoma medication.
Since each study used a different wash-out protocol, when post-wash-out IOP was measured at baseline and in the various follow-ups, these data were used in the analysis; when impossible, IOP values during pharmacological treatment were used instead.
Data were extracted manually from the chosen studies.All selected papers underwent a quality verification process using the SIGN checklist (Scottish Intercollegiate Guidelines Network).

Analysis and Data Synthesis
A meta-analysis is a quantitative technique allowing, from a clinical and statistical standpoint, for researchers to combine different studies on a single clinical issue to reach a single conclusive result with a more consistent statistical power.
The meta-analysis is completed through Statsdirect 3.2 (StaTsdirect, Wirral, UK).The data we focus on are average IOP pre and post surgery and their standard deviation.In the studies where the confidence interval is not reported, the standard error and/or p-values are extracted to calculate average and standard error of the difference compared to baseline (t0).
Using a random effects model, we can attempt to estimate the average of the true effect's distribution.Bigger studies can then lead to more accurate estimates compared to smaller studies, but all effect sizes are included when estimating the average.The assigned weights under a random effects model are more balanced than under the fixed effects model, as larger studies are less likely to dominate the analysis and smaller studies are less likely to be overshadowed.
The pooled difference allows us estimation of the common difference by assuming that all different populations have the same variance.The Z test enables us to evaluate a hypothesis if the population variance is known or the sample size is ≥30.
Non-combinability is checked through the Cochran Q test to verify whether the treatments have the same effect.Heterogeneity is calculated through the I 2 test and is significant for values between 75% and 100%, working in favour of analysing the metaanalysis through the random effects model.
The DerSimonian and Laird random effects model has a starting point of considering the effects of the studies as different but related.It is chosen because, by accounting for the variance in effect size, it allows us the use of this information and the results to make inferences on how operating to reduce IOP can benefit other glaucoma patients, as per the hypothesis.
The bias indicators used are the Begg and Mazundar rank correlation, a test estimating the correlation between the degrees of the effect size and the degrees of its variance, as well as the Egger regression, which outputs the degree of asymmetry in the funnel plot graphically represented in the Bias Assessment Plot.
Data summarised in the meta-analysis are graphically reported via the random effects model through a forest plot.

Results
Table 2 collects all studies satisfying the inclusion/exclusion criteria from the search of the databases with their respective average differences compared to baseline with relative standard deviation at 6, 12, and 24 months.The following results emerged from the analyses of the four sub-groups (POAG (Table A1), ACG (Table A2), PXG (Table A3) and OH (Table A4)).
The results show a significant reduction in IOP after cataract extraction surgery at all three time-points of 6, 12, and 24 months for POAG.Given the high heterogeneity of the included studies in this subgroup at any timepoint (respectively, I 2 = 78%, I 2 = 94% and I 2 = 94.7%), in the analysis of IOP reduction in POAG, a random effects model (DerSimonian & Laird) was applied.The average combined difference was 2.44 mmHg at 6 months, 2.71 mmHg at 12 months, and 3.13 mmHg at 24 months (95% CI).The Z-tests of the differences all had p-values of less than 0.0001, indicating significance (Figure 2).
The forest plots demonstrate that the data sets are of good quality and have no publication bias, as evidenced by the symmetry in the funnel plots (Figure A1).The grey boxes in the forest plots represent the effect size of single studies, and the grey rhombus which represents a combined difference greater than one indicates a significant association.The same holds true, at identical timepoints, for angle closure glaucoma.The analysis shows that, after phacoemulsification, there is a significant reduction in IOP in ACG patients at 6, 12 and 24 months from baseline (Figure 2).The results are based on nine included studies for the 6-and 12-month analysis and eight included studies for the 24-month analysis.The high heterogeneity of the included studies was, once again, accounted for by using the DerSimonian and Laird Random Effects Model.The average combined difference in IOP reduction was 6.81 mmHg (95% CI = 4.06 to 9.55) at 6 months, 7.03 mmHg (95% CI = 4.26 to 9.81) at 12 months and 6.52 mmHg (95% CI = 3.84 to 9.21) at 24 months.The Z-test results indicated that the reduction, with a p-value of less than 0.0001, was significant (Figure 3).
The forest plot and funnel plot (Figure A2) analysis also showed that the ACG data set is of good quality, and there was no evidence of publication bias, as confirmed by the Begg and Mazundar rank correlation and Egger regression results.
Despite the scarcity of eligible studies, even in the two subgroups, PXG and OH, the results showed that the reduction in IOP was significant at 12 months from baseline in PXG with a combined difference of 5.30 mmHg (95% CI = 2.216671 to 8.375508), and significant at 24 months from baseline in OH, with a combined difference of 2.27 mmHg (95% CI = 0.106467 to 4.444148) (Figure 4).
The random effects model was also applied in these subgroups due to the high heterogeneity, and the forest plot and funnel plot (Figure A3) suggest that no publication bias was detected.Forest plots meta-analysis on PXG at 12 months and on OH at 24 months from surgery [15][16][17][18]22,29,36,37].

Discussion
The systematic literature review and subsequent meta-analysis were executed to analyse data already available in the literature relating to the isolated procedure of cataract extraction with an IOL implant on pre-surgical IOP in patients with POAG, ACG, PXG and OH.
As evident in Figure 5, the highest reduction in IOP, in each of the analysed t x , is always in Angle-Closure Glaucoma, with values of 6.81 mmHg, 7.03 mmHg e 6.52 mmHg.Inversely, the lowest reduction, in absolute terms, is that of Ocular Hypertension, with a value of 2.28 mmHg at 24 months from surgery.

Discussion
The systematic literature review and subsequent meta-analysis were executed to analyse data already available in the literature relating to the isolated procedure of cataract extraction with an IOL implant on pre-surgical IOP in patients with POAG, ACG, PXG and OH.
As evident in Figure 5, the highest reduction in IOP, in each of the analysed tx, is always in Angle-Closure Glaucoma, with values of 6.81 mmHg, 7.03 mmHg e 6.52 mmHg.Inversely, the lowest reduction, in absolute terms, is that of Ocular Hypertension, with a value of 2.28 mmHg at 24 months from surgery.Various theories were hypothesised in trying to explain these reductions in ocular hypertension:

•
The molecular theory based on the effects on the pattern of the trabecular meshwork: the inflammatory reaction, consequent from surgery, could lead to hyposecretion of aqueous humour, a reduction in resistance to outflow and biochemical alterations in the blood-aqueous barrier.

•
The physiologic theory based on the effects on the ciliary body: it appears that cataract extraction has a relevant effect on the dynamic involving the ciliary body by reducing its anteposition, especially relevant in ACG.

•
The biomechanical theory based on the anatomical changes in the anterior segment: with an improvement in predictive anatomical parameters on the reduction in IOP in an OCT scan, mainly in the aperture of the camerular angle; The biomechanical theory based on the position of the lens: since an excessively anterior position favours the formation of a higher pressure gradient, which can lead to relative pupillary block.

•
The biomechanical theory based on fluid dynamics: the high flow generated from phacoemulsification, in this limited anatomical space, can clean the pattern of the trabecular meshwork and favour the action of the macrophages in that location [38].
Numerous studies have now shown that cataract extraction, significant in the visual field, can lead to an improvement in sight with an improvement in associated quality of life, especially in patients with pre-existing damage to the visual field, e.g., glaucoma patients.These benefits need to be attentively discussed with glaucoma patients, explaining accurately the involved risks and the implications of surgery [39].Various theories were hypothesised in trying to explain these reductions in ocular hypertension:

•
The molecular theory based on the effects on the pattern of the trabecular meshwork: the inflammatory reaction, consequent from surgery, could lead to hyposecretion of aqueous humour, a reduction in resistance to outflow and biochemical alterations in the blood-aqueous barrier.

•
The physiologic theory based on the effects on the ciliary body: it appears that cataract extraction has a relevant effect on the dynamic involving the ciliary body by reducing its anteposition, especially relevant in ACG.

•
The biomechanical theory based on the anatomical changes in the anterior segment: with an improvement in predictive anatomical parameters on the reduction in IOP in an OCT scan, mainly in the aperture of the camerular angle;

•
The biomechanical theory based on the position of the lens: since an excessively anterior position favours the formation of a higher pressure gradient, which can lead to relative pupillary block.

•
The biomechanical theory based on fluid dynamics: the high flow generated from phacoemulsification, in this limited anatomical space, can clean the pattern of the trabecular meshwork and favour the action of the macrophages in that location [38].
Numerous studies have now shown that cataract extraction, significant in the visual field, can lead to an improvement in sight with an improvement in associated quality of life, especially in patients with pre-existing damage to the visual field, e.g., glaucoma patients.These benefits need to be attentively discussed with glaucoma patients, explaining accurately the involved risks and the implications of surgery [39].
The correlation between cataract surgery and glaucoma comprises numerous facets but this only helps in understanding the new possible role that this surgical procedure has in terms of addressing glaucoma.Cataract extraction alone can help in reducing IOP, in selected patients, trying to improve their visual ability, in the short and long term [38].
Preceding meta-analyses, such as Masis Solano et al. from 2018 [40], have already found significant IOP reduction compared to baseline at the end of the follow-ups: 2.7 mmHg for POAG, 6.4 mmHg for ACG, and 5.8 mmHg for PXG.
With this paper, we instead wanted to offer an indication of how much IOP is lowered over time, to help ophthalmologists understand whether, with cataract surgery alone, it is possible to have a reduction such as to reach the target IOP at a pre-established time interval and to maintain it in the medium term.
It is notable that in the Masis Solano study [40], despite the commonly held idea that phacoemulsification on its own is not an effective POAG treatment, it was already stated that, albeit modest compared to IOP reduction in ACG, cataract extraction is a possible alternative in lighter cases in which the safety of the procedure is the main concern.In cases of ACG, the reduction is significant enough and the operation can be considered a first course of action [41].
As reported in the Ocular Hypertension Treatment Study, this surgical procedure safely reduces IOP even in cases of simple ocular hypertension, but it cannot be confidently said that it also reduces the risk of developing glaucoma [16].
Although this analysis has significant results, its limitations must be acknowledged.Data of studies that contemplated therapeutic washout, often with different protocols in performing it, and of other studies in which this was not performed were aggregated.Further work comparing the former with the latter would certainly be interesting to evaluate the weight of this protocol on the effect of cataract surgery in managing IOP, when not influenced by pharmacological therapy.
Among the variables to consider, the surgical technique for cataract extraction deserves further investigation, as phacoemulsification is not available in all areas of the world.However, as reported by Sengupta et al., it appears that the reduction in IOP using Manual Small Incision Cataract Surgery (MSICS) is comparable at 6 months post surgery [42].
Another aspect that certainly deserves evaluation is how treatment is modified over time after phacoemulsification, drawing on this to better understand the extent of the individual active ingredients' effect in the post-operative period.

Conclusions
Although additional research is needed to delve into the individual mechanisms and variables of IOP reduction, there are benefits of cataract phacoemulsification in patients with glaucoma and ocular hypertension, not only shortly after surgery, but also in the following years and over the long term in managing IOP.
The effect is surely striking in Angle-Closure Glaucoma but is not to be underestimated in Primary Open Angle Glaucoma and in its Pseudo-Exfoliative subtype, where it is even more apparent.
It is also important to note how, even in patients with just Ocular Hypertension, a non-insignificant reduction in IOP can be found, useful in protecting from the damage it could cause if it were to evolve into a glaucoma even though we cannot be sure to which degree the eventual perimetrical damage evolution would be affected by this surgery, neither in OH nor in manifest glaucoma.
The starting point of this paper was demonstrating the correlation between cataract surgery and the main modifiable factor in glaucoma, intraocular pressure.
Having shown the existence of this correlation, with phacoemulsification reducing IOP, as well as its statistical significance in all subgroups, we suggest inserting this procedure in the therapeutic framework for other subgroups, as it is already the case for ACG.We do nonetheless fully acknowledge the risks, potentially even catastrophic, it presents, but our findings suggest it should be considered to help reach the target IOP value for a patient's eye.

Figure 5 .
Figure 5. Bar chart of average IOP reduction in mmHg in the 4 subgroups at 6, 12 and 24 months.

Figure 5 .
Figure 5. Bar chart of average IOP reduction in mmHg in the 4 subgroups at 6, 12 and 24 months.

Figure A1 .
Figure A1.Funnel plots for bias assessment with corresponding indicators in the included studies of POAG subgroup at 6, 12 and 24 months.Figure A1.Funnel plots for bias assessment with corresponding indicators in the included studies of POAG subgroup at 6, 12 and 24 months.

Figure A1 .
Figure A1.Funnel plots for bias assessment with corresponding indicators in the included studies of POAG subgroup at 6, 12 and 24 months.Figure A1.Funnel plots for bias assessment with corresponding indicators in the included studies of POAG subgroup at 6, 12 and 24 months.

Figure A2 .
Figure A2.Funnel plots for bias assessment with corresponding indicators in the included studies of ACG subgroup at 6, 12 and 24 months.Figure A2.Funnel plots for bias assessment with corresponding indicators in the included studies of ACG subgroup at 6, 12 and 24 months.

Figure A2 .
Figure A2.Funnel plots for bias assessment with corresponding indicators in the included studies of ACG subgroup at 6, 12 and 24 months.Figure A2.Funnel plots for bias assessment with corresponding indicators in the included studies of ACG subgroup at 6, 12 and 24 months.

Figure A3 .
Figure A3.Funnel plot for bias assessment with corresponding indicators in the included studies of PXG subgroup at 12 months and of OH subgroup at 24 months.Figure A3.Funnel plot for bias assessment with corresponding indicators in the included studies of PXG subgroup at 12 months and of OH subgroup at 24 months.

Figure A3 .
Figure A3.Funnel plot for bias assessment with corresponding indicators in the included studies of PXG subgroup at 12 months and of OH subgroup at 24 months.Figure A3.Funnel plot for bias assessment with corresponding indicators in the included studies of PXG subgroup at 12 months and of OH subgroup at 24 months.

Table 1 .
List of studies satisfying the inclusion and exclusion criteria and their characteristics.

Table 2 .
Author, year of publication, number of eyes, average difference with relative SD at 6, 12, and 24 months, study type and glaucoma subtype for the chosen studies as presented from the authors.

Table A2 .
Standardised effect, standard error and study weight in ACG respectively at 6, 12 and 24 months from surgery.

Table A3 .
Standardised effect, standard error and study weight in PXG at 12 months from surgery.

Table A4 .
Standardised effect, standard error and study weight in OH at 24 months from surgery.