Effect of cataract surgery on wet age‐related macular degeneration activity

Wet age‐related macular degeneration (AMD) and age‐related cataract are often coexisting causes of visual impairment. Yet, the timing of cataract surgery in wet AMD patients is controversial.


Introduction
Age-related macular degeneration (AMD) is the leading cause of visual impairment in western countries with increasing global incidence (Wong et al. 2014;Mitchell et al. 2018). Age-related cataracts often coexist with AMD (Ho et al. 2008), with reported comorbidity in 74% of visually impaired patients over 65 years of age (Kessel et al. 2015;Flaxman et al. 2017). Wet AMD can cause vision to deteriorate rapidly if left untreated. In recent years, antivascular endothelial growth factor (anti-VEGF) injections have revolutionized the prognosis of wet AMD.
The timing of cataract surgery among wet AMD patients is a matter of debate (Kessel et al. 2015). Apparently, it might be reasonable to postpone cataract surgery when there is concern about wet AMD exacerbation, as cataract surgery might induce an inflammatory response. It is postulated that surgery may disrupt or further stimulate delicate leaking neovessels characteristic of wet AMD. Consequently, surgery might stimulate progression of geographic atrophy, development of choroidal neovascularization (CNV) or lead to a weaker anti-VEGF treatment response (Saraf et al. 2015). On the other hand, cataract reduces retinal visibility and hinders treatment, in addition to reducing patients' vision. It is, therefore, not completely clear whether wet AMD should be treated with the normal anti-VEGF protocol despite cataract surgery, and whether it is justified to postpone cataract surgery depending on disease activity.
Comorbidities, especially diabetes, and patients' systemic medication such as vasoactive drugs were suggested to have a bearing on the incidence of pseudophakic cystoid macular oedema (PCME) after routine cataract surgery (Ylinen et al. 2018;Danni et al. 2019). Antihypertensive drugs and statins have not been found to be significantly associated with the progression of AMD (Al-Holou et al. 2015;Heesterbeek et al. 2020), while anticoagulation and antiplatelet drugs have been considered to carry at least a theoretical risk of exacerbating wet AMD (Small et al. 2017).
Several studies have evaluated the influence of cataract extraction on AMD progression (Cugati et al. 2006;Chew et al. 2009). Nevertheless, there is a lack of consensus regarding the protocol treatment of wet AMD patients undergoing cataract surgery (Othenin-Girard et al. 1994;Chew et al. 2009;Kessel et al. 2014;Ylinen et al. 2018). A better understanding of the variables associated with macular anatomical and functional response to treatment with anti-VEGF in wet AMD patients and to cataract surgery could improve preoperative planning and reduce the treatment burden. Here, we aim to investigate whether cataract surgery in wet AMD patients affects its course and the relationship to how it influences clinical outcomes.

Study design
This study was a registry-based retrospective cohort study of consecutive cataract surgeries performed on wet AMD patients at Helsinki University Hospital, Helsinki, Finland. The study was approved by the Helsinki and Uusimaa Hospital District (Research permission §42/2019; HUS/466/2019) and complied with the tenets of the Declaration of Helsinki.
The primary outcome of the study was to evaluate the change in CSMT after surgery and at 1 year. Secondary outcomes were the changes in visual acuity, anti-VEGF treatment interval, macular status between the preoperative phase and postoperatively and the correlation between the CSMT change and the cumulative number of anti-VEGF injections before surgery.

Data acquisition
The study included patients who underwent uncomplicated cataract surgery between January 2014 and December 2018 at Helsinki University Hospital. Data were collected from the operating room management system (BCB Medical, cataract database, Turku, Finland). 25,818 consecutive cases of uncomplicated cataract surgery were reviewed for registry. For inclusion, all patients were required to have documentation on the type of postoperative medical treatment received, systemic medication and anti-VEGF treatment for wet AMD prior to laterality matched uncomplicated cataract surgery.

Anti-VEGF treatment protocol
Helsinki University Hospital uses the fixed-treat-and-extend regimen (fixed-TER) anti-VEGF protocol for the treatment of wet AMD. Either bevacizumab 1.25mg (Avastin ® , Genentech, Inc., South San Francisco, CA) or aflibercept 2mg (Eylea ® , 40mg/ml, Bayer Pharma AG, Leverkusen, Germany) was used as anti-VEGF drug according to the national current care guidelines for wet AMD (Tuuminen et al. 2017).
Patients received three monthly anti-VEGF injections at baseline, and the treatment interval was lengthened accordingly: three consecutive injections at 6-week interval, two consecutive injections at 8-week interval, one injection at 10-week interval and finally reaching the maximum 12-week interval. If the disease showed activation, the treatment intervals were shortened.

Cataract surgery
According to the Finnish national guidelines for cataract operations, the best-corrected visual acuity (BCVA) for cataract surgery is recommended to be 0.5/0.3 in the better/worse eye, respectively, or less by Snellen equivalents, except under some specific circumstances (Cataracts Current Care Guideline 2019). The surgical technique used in this study was phacoemulsification (Infinity ® /Centurion ® Vision System, Alcon Laboratories, Inc., Fort Worth, TX) with a 2.4-2.75 mm clear corneal incision. A single-piece acrylic monofocal IOL (Tecnis ® , Johnson & Johnson Surgical Vision Inc., Santa Ana, CA or AcrySof IQ ® , Alcon Laboratories, Inc.) was implanted into the capsular bag. During the study period, the single-piece monofocal Tecnis ® was the IOL type used most (96.1%) (Hecht et al. 2020). None of the wet AMD patients underwent complicated cataract surgery.

Postoperative care
The choice of postoperative anti-inflammatory medication was at the discretion of the treating physician. Treatment was either with steroids (N = 22, 19.8%; Pred Forte ® , prednisolone acetate 10 mg/ml, Allergan, Inc. Irvine, CA, or Oftan Dexa ® , dexamethasone 1 mg/ml, Santen Oy, Tampere, Finland), NSAIDs (N = 80, 72.1%; Nevanac ® , nepafenac 1mg/ml or 3mg/ml, Novartis, Basel, Switzerland) or a combination of both (N = 9, 8.1%). The duration of the anti-inflammatory treatment was three weeks in all cases except for two in which the surgeries were planned with four-week anti-inflammatory treatment. No subconjunctival or intracameral anti-inflammatory adjunct treatment was used.

Clinical evaluation
All patients underwent a comprehensive ophthalmic examination and optical coherence tomography (OCT) analysis. Foveal thickness, central subfield macular thickness (CSMT) defined as mean thickness in the central 1000 μm diameter area and maximum thickness in the central 1000 μm diameter area were recorded by spectraldomain OCT (SD-OCT, Heidelberg Eye Explorer Version 1.10.4.0 and HRA / SPECTRALIS ® Viewing Module Version 6.13.3.0, Heidelberg Engineering GmbH, Heidelberg, Germany). Follow-up 30-frame SD-OCT scans were performed with AutoRescan TM software. The subtyping of wet AMD was done at the time of diagnosis with fluorescein angiography (FA) and/or indocyanine green angiography. Macular status on OCT scans of all cases was evaluated masked by a medical retina specialist (P.K.).
Visual acuity was examined using a standard Snellen chart with the bestcorrected refraction and using decimal numbers. Best-corrected visual acuity (BCVA) decimal values were converted to logarithms of the minimum angle of resolution (logMAR) for statistical purposes. BCVA and OCT parameters were obtained at different time points, including: at the time of wet AMD diagnosis, and within 18 weeks before and after surgery depending on the timing of out-patient clinic visits according to our fixed-TER protocol and patients' treatment intervals, that is anti-VEGF injections may have been given in between the surgery and the latest OCT scans.

Sample size calculation
For sample size calculations, the standard deviation of CSMT was set at 70 μm (according to previous studies ranging from 61 to 80 μm) (Berg et al. 2015;Achiron et al. 2021) with a minimal clinically significant difference of 10 μm in CSMT. Using a paired model, we found that a sample size of 98 patients was required to detect a significant difference between measurements with a significance level of 0.05 and a power of 80%.

Statistics
Unless otherwise specified, data are presented as mean AE standard deviation (SD). Visual acuity decimal numbers were converted to logMAR units using the -log10 conversion method. In visual acuity testing, when the largest optotype could not be recognized correctly, the classification was very low VA on a semi-quantitative scale and was graded as counting fingers (CF) and hand motion (HM). Very low VA measurements were converted as follows: CF to 1.9 and HM to 2.3 logMAR units (Dong et al. 2003;Schulze-Bonsel et al. 2006). Statistical analysis was performed using IBM SPSS Statistics 25 (SPSS Inc., Chicago, IL). For two-group comparison, the chi-squared test was used for categorical variables, Student's t-test was used for parametric variables and the Mann-Whitney U-test for non-parametric variables. The paired t-test was used to compare BCVA, CSMT and anti-VEGF treatment interval at different timepoints. For multiple group comparisons, the Fisher-Freeman--Halton test was used for qualitative data, the one-way ANOVA test using Bonferroni's correction was used for continuous variables, and the Kruskal-Wallis test with Dunn's correction for non-parametric variables. A multiple linear regression model was used to estimate the relationships between the preoperative CSMT, cumulative number of anti-VEGF injections, anti-VEGF treatment interval, macular status (presence of subretinal and intraretinal fluid) at cataract surgery and the CSMT change postoperatively. p < 0.05 was considered statistically significant.

Results
One hundred and eleven eyes of 111 patients (38 males and 73 females) with a mean age of 78.9 AE 5.6 years were included in the study (Table 1) Table 3). After cataract surgery, foveal thickness increased >30% (from preoperative values) in only three out of the 111 eyes (2.7%), a CSMT increase of >30% was found in one case (total incidence 0.9%) and a maximal macular thickness increase >30% in one case (total incidence 0.9%) ( Table 3). Postoperatively, none of the patients lost BCVA of 0.2 Snellen decimals or more compared to preoperative values (data not shown). At 1 year, two patients lost BCVA of 0.2 Snellen decimals or more compared to their preoperative values (data not shown). The anti-VEGF treatment interval was 6.53 AE 2.08 weeks prior to surgery, 7.03 AE 2.23 weeks at 1 year (p = 0.246, Table 3) and 7.05 AE 2.57 weeks at the last documented visit (p = 0.035, Table 3).
The CSMT change postoperatively inversely correlated with the CSMT level prior to surgery (i.e. the higher the macular thickness preoperatively the higher the CSMT resolution postoperatively, r = −0.483, p < 0.001, Fig. 1), but not with the cumulative number of anti-VEGF injections before surgery (r = −0.051, p = 0.691, Fig. 2). The CSMT change postoperatively or at 1 year did not correlate with patient age at surgery, nor with the preoperative BCVA (Table 4).
In a multiple regression analysis, significant baseline variables such as preoperative CSMT, cumulative num- Clinical outcomes after cataract surgery regarding different wet AMD subtypes and the presence of intraretinal or subretinal fluid during surgery are presented in Tables S1 and 5, respectively.
Effect of systemic drugs and topical antiinflammatory medication on cataract surgery outcomes in eyes with wet AMD To assess the influence of baseline factors such as existence and type of diabetes, anti-inflammatory drugs, use of anti-coagulants, calcium-channel blockers and statins on the study  The effect of baseline factors was also examined independently for an association with clinical outcomes using a multiple regression analysis. Baseline factors included existence and type of diabetes, systemic drugs (anticoagulants, calcium-channel blockers and statins) and topical anti-inflammatory medication along with age and gender, and were examined for association with CSMT change postoperatively or at 1 year and the anti-VEGF treatment burden after surgery. None of the baseline factors were associated with any clinical outcomes related to wet AMD first or 1-year postoperative visits, except for thinner CSMT at first postoperative visit in those taking statins (B −23.17, 95% CI −41.1 to −5.24, p = 0.012).

Discussion
In this retrospective registry-based study, we show that cataract surgery does not cause progression or worsening of wet AMD. Among the 111 eyes investigated only one eye showed an increase of over 30% in CSMT and in maximal macular thickness. The mean treatment interval of anti-VEGF injections did not change after surgery compared to the preoperative period. In addition, BCVA improved significantly up to 1 year after surgery.
Cataract and wet AMD often coexist, and both diminish visual function. Following cataract surgery, wet AMD can theoretically progress through one of two possible mechanisms: blue-light toxicity and surgically induced inflammation. Blue light has been shown in laboratory studies to affect retinal pigment cells (Noell et al. 1966;Ham et al. 1976;Grimm et al. 2001;Algvere et al. 2006) and is thought to have an in vivo negative effect on macular health (Sparrow et al. 2002). This led to the invention of blue-light filtering intraocular lenses (IOLs). In the present study, only non-blue-light filtering IOLs were implanted; nevertheless, we still could not show wet AMD progression after cataract surgery. Indeed, in a recent study by our group, nonblue-light filtering IOLs were ruled out as a possible explanation for wet AMD progression, as they were found equal to blue-light filtering IOLs in terms of the development and progression of wet AMD (Achiron et al. 2021).   Regarding surgically induced inflammation, the stress to the eye with modern phacoemulsification technology is much less than with older methods such as intracapsular and extracapsular cataract extraction. Consequently, PCME and other intraand postoperative complications are fewer with phacoemulsification surgery (de Silva et al. 2014). This has been partially attributed to the use of anti-inflammatory medications in the immediate postoperative period (Kessel et al. 2014;Hecht et al. 2020). For most patients in the present study, the postoperative anti-inflammatory treatment period after cataract surgery was three weeks. Neither topical anti-inflammatory medication administered during the postoperative period nor the use of systemic medication was associated with macular status. The only significant measure found was that the thicker the CSMT at the time of surgery and the greater the extent of CSMT resolution postoperatively, the better was the BCVA gain. In the vast majority of patients, anti-inflammatory treatment was given three weeks after cataract surgery. Since events of wet AMD activation and clinically significant relapses were rare, we speculate that three weeks of antiinflammatory medication is sufficient, regardless of the presence of wet AMD. Of note, the anti-inflammatory treatment and the incidences between NSAID, corticosteroids and their combination in wet AMD patients reflect the real-life of clinical practice, recently studied by us in another article (Hecht et al. 2020).
We also evaluated the effect of type II diabetes mellitus on wet AMD progression and found no effect on BCVA, CSMT or the number of anti-VEGF injections after cataract operation. Moreover, in parallel with the age-related eye disease (AREDS) report (Al-Holou et al. 2015), statins were not associated with wet AMD progression in our study. Similarly, anticoagulation and other systemic medication such as CCB and betablockers were not associated with wet AMD progression either. Saraf et al. (2015) have evaluated the effect of cataract surgery on wet AMD progression in a case-control study. As in our study, they demonstrated no change in the number of anti-VEGF injections before and after cataract surgery. Interestingly, they reported an increase in the mean central thickness and worsening of foveal cysts during the 3 months after cataract surgery. They postulated that these OCT changes may be related to PCME occurrence in the early postoperative period. However, the OCT features were not followed beyond the 3-month postoperative period. Our data did not indicate worsening of OCT findings within the postoperative time points. Furthermore, OCT findings remained stable even after a longer period of follow-up of up to 1 year.
When considering clinically relevant parameters and the best proxy for wet AMD activity the clinical variables CSMT, visual acuity, macular status and anti-VEGF injection interval could be considered as candidates. Visual acuity and its gain during the postoperative phase is largely dependent on the baseline status of the macula and other comorbidities, rather than on wet AMD activity alone. Macular status (presence or absence of fluid) is a dichotomic variable that does not reflect the severity of the exudation. Anti-VEGF treatment interval is gradually changing towards a shorter or a longer time, and thus better reflects the long-term rather than the current activity of the disease. On the other hand, of the clinically relevant variables, CSMT change was considered to best correlate with the current wet AMD activity. Thus, CSMT change was selected as a primary outcome. Furthermore, we investigated which baseline variables Data are given as mean AE SD or absolute numbers. Data were analysed with the Student's t-test for continuous variables and with Mann-Whitney U-test for non-parametric variables. '+' sign represents the presence and '−' sign represents the absence of a factor in question. BCVA = best-corrected visual acuity, CSMT = central subfield macular thickness, IRF = intraretinal fluid, logMAR = logarithm of the minimum angle of resolution, SRF = subretinal fluid.
correlate with CSMT change postoperatively and at 1 year. We found a significant inverse correlation with baseline CSMT. These results can be interpreted to mean that the higher the CSMT at baseline, the greater the extent of CSMT resolution at the postoperative timepoints. Further, evidenced by the resolution of exudation regardless of cataract surgery, these results can be interpreted to indicate that the wet AMD activity and the large extent of exudation in wet AMD patients prior to cataract surgery do not indicate the need to postpone the surgery.
According to a recent survey, 12% of ophthalmic surgeons in the UK did not consider cataract surgery for wet AMD patients due to fear of disease activation (Malick et al. 2020). Other recent studies suggested postponing cataract surgery for up to 6 months from wet AMD diagnosis because of ongoing disease activity (Kessel et al. 2015;Kessel et al. 2016;Daien et al. 2018). On the other hand, some studies concluded that cataract surgery does not cause or worsen the course of wet AMD and that surgery may be performed at any time despite the presence of macular fluid detected by OCT (Wang et al. 2016;Ehmann & Ho 2017;Rim et al. 2017;Starr et al. 2018). A recent review concluded that even if cataract surgery does affect AMD, this hazard is of insignificant magnitude relative to the established role of ageing as an AMD risk factor (Ehmann & Ho 2017). This small risk would be far outweighed by the benefits of improved visual function and a better quality of life after surgery. Similarly, our study supports performing cataract surgery regardless of wet AMD activity when a potential improvement in visual acuity can be expected. Neither the macular status prior to surgery nor the cumulative number of anti-VEGF injections was associated with clinical outcomes postoperatively. Helsinki University Hospital has had no protocol to administer anti-VEGF just before or at the end of cataract surgery. Only two surgeries out of 111 were performed with concomitant anti-VEGF administration. In the light of our findings, it may be unnecessary to time anti-VEGF injection just before or at the same time as cataract surgery, but this is beyond the scope of this study and deserves to be studied in a separate prospective trial.
A relatively small cohort size (N = 111) and the short follow-up time are the primary limitations of this study. Although 1 year of follow-up is relatively short, it is reasonably long enough to assess CSMT, visual acuity gain and anti-VEGF burden. Our study was also designed to address the question of whether postoperative management of cataract surgery should be standardized or tailored individually in wet AMD. We found that the incidence of a clinically significant CSMT increase (0.9%) as defined by a 30% CSMT increase was comparable to that previously reported among cataract surgery patients without known risk factors for PCME (1.17%) (Chu et al. 2016). Furthermore, a postoperative trend towards a CSMT decrease was achieved in wet AMD patients, even after cataract surgery regardless of disease activity. The anatomical and functional outcomes of surgery were not dependent on anti-VEGF treatment status or activity of wet AMD. Neither topical anti-inflammatory drug selection nor systemic medication was associated with positive the long-term outcome of cataract surgery. In addition, previous studies have shown that wet AMD patients' motor function and visionrelated quality of life both improve after cataract surgery (Taipale et al. 2020). On the other hand, postponing cataract surgery due to coexisting wet AMD may adversely affect health-related quality of life (Porela-Tiihonen et al. 2016) and can be an obstacle to treatment due to poor retinal visibility (Loriaut et al. 2014). Another limitation of this study is that the time between pre-and postoperative outpatient visits and cataract surgery was not standardized because of the retrospective setting, which means that the macular oedema may fluctuate depending on the cataract surgery and treatment intervals. On the other hand, use of just a single standardized timepoint, such as a postoperative visit four weeks after surgery, would lack the information on macular status just before the next anti-VEGF administration. In TER protocol all out-patient visits were predetermined based on the disease activity and the current treatment interval. Thus, we find that the macular status at the predetermined out-patient visit is of clinical interest.
We conclude that timing of cataract surgery in wet AMD patients should be adjusted according to patients' needs. We found no justification to support delaying surgery until dry macula has been achieved. Surgery should, therefore, not be postponed for these considerations in patients who require it.