Long-Term Variations in Retinal Parameters after Uncomplicated Cataract Surgery

Background: Cataract phacoemulsification surgery provides excellent refractive results; however, it also elicits changes in the posterior segment of the eye. This study aimed to determine changes in retinal parameters measured by spectral-domain optical coherence tomography (SD-OCT) and OCT angiography (OCTA) after an uncomplicated cataract surgery, including the impact of effective phacoemulsification time (EPT). Methods: The study included 44 patients without retinal abnormalities, followed up after unilateral uncomplicated cataract phacoemulsification in a single ophthalmological unit. Patients were evaluated for the following parameters at baseline and at 2 weeks, 3 months, and 12 months after the surgery: best corrected visual acuity, central retinal thickness (CRT), average central retinal thickness (CRTA), central retinal volume (cube volume (CV)), vessel density central (VDC), vessel density full (VDF), vessel perfusion central (VPC), and vessel perfusion full (VPF). The EPT recorded at each procedure was used as a covariant for the evaluation of changes in retinal parameters after the surgery. Analysis included 44 eyes for SD-OCT and 17 for OCTA evaluation, according to adopted scan quality thresholds. Results: A significant increase in CRT, CRTA, and CV was noted at each follow-up point compared with baseline. The rising tendency was observed in the first 3 months after the surgery, with a decline over the subsequent months. The VPF parameter showed a stable improvement after the surgery. The analysis of covariance did not confirm any significant effect of the EPT on variations in CRT, CV, CRTA, VDC, and VPF and there was a weak effect on the VDF parameter. Conclusions: Uncomplicated cataract surgery results in an increase in retinal thickness and volume in the first few months after the surgery, followed by a spontaneous decline in these parameters in the subsequent months. A long-standing improvement is noted in the VPF parameter.


Introduction
Cataract surgery is a common routine procedure in ophthalmology. Its effects are usually evaluated in the context of postoperative refraction and visual improvement. However, even uncomplicated cataract extraction may alter the normal retina by inducing an inflammatory process, functional hyperemia, and changes in ocular perfusion pressure [1][2][3][4][5][6]. Information on the intensity and persistence of these alterations could be valid in the context of possible functional improvements but also potential risks (such as the development of retinal disorders).
Variations in retinal thickness after an uncomplicated cataract surgery have been already assessed in the ophthalmological literature [7][8][9][10][11][12][13][14][15][16][17][18][19]. Most studies used optical coherence tomography (OCT) to detect alterations in retinal thickness and the incidence of cystoid macular edema. The advent of OCT angiography (OCTA) has made it possible BCVA was tested at each visit on the Snellen chart, and the results were subsequently converted to logMAR units. The applied correction was based on the automated refraction measured with the automated refractor (Huvitz HRK-1, Dongan-gu, Republic of Korea, 2018).
SD-OCT examination was performed using the Zeiss AngioPlex SD-OCT (Carl Zeiss Meditec AG, Jena, Germany), which enables the OCT angiography measurements of vessel density and perfusion. In this particular device, the Early Treatment of Diabetic Retinopathy grid is used to determine the measured areas of the central retina. CRT refers to the mean retinal thickness within the central part of the macula of 1 mm in diameter. CRTA corresponds to the mean retinal thickness of the central retinal area of 6 mm in diameter. CV reflects the retinal volume within the central retinal circle of 6 mm in diameter, which approximately refers to the macular area. Vessel density and perfusion density are parameters introduced in the AngioPlex device for the vasculature of the retinal superficial capillary plexus. The scans are obtained from the central retinal square area of 3 mm × 3 mm. Vessel density is defined as the total length of perfused vasculature per unit area and expressed as mm/m 2 . On the other hand, perfusion density measures the percent total area of perfused vasculature in a given region of measurement. In addition to the length, it also involves the caliber of the vessels. Perfusion density is a relation of the number of pixels with perfused vasculature to the total number of pixels in the region; thus, it is expressed in %. Central vessel and perfusion densities (VDC and VPC, respectively) represent the parameters in the central circle of 1 mm in diameter. On the other hand, full vessel and perfusion densities (VDF and VPF, respectively) refer to the central area of 3 mm in diameter.
The study was approved by a local bioethics committee (Komisja Bioetyczna at OIL at OIL in Gdańsk; approval no.: KB-18/20 dated 7 July 2020).

Statistical Analysis
Statistical analysis was performed using the IBM SPSS Statistics 27 software. (IBM, Armonk, NY, USA) The effects of cataract surgery on retinal parameters were assessed by the repeated measures analysis of variance (ANOVA) for the four time points with the inclusion of the EPT covariant (ANCOVA). The analysis included the following parameters: BCVA, CRT, CRTA, and CV. Each of these parameters was assessed with the same SD-OCT examination was performed using the Zeiss AngioPlex SD-OCT (Carl Zeiss Meditec AG, Jena, Germany), which enables the OCT angiography measurements of vessel density and perfusion. In this particular device, the Early Treatment of Diabetic Retinopathy grid is used to determine the measured areas of the central retina. CRT refers to the mean retinal thickness within the central part of the macula of 1 mm in diameter. CRTA corresponds to the mean retinal thickness of the central retinal area of 6 mm in diameter. CV reflects the retinal volume within the central retinal circle of 6 mm in diameter, which approximately refers to the macular area. Vessel density and perfusion density are parameters introduced in the AngioPlex device for the vasculature of the retinal superficial capillary plexus. The scans are obtained from the central retinal square area of 3 mm × 3 mm. Vessel density is defined as the total length of perfused vasculature per unit area and expressed as mm/m 2 . On the other hand, perfusion density measures the percent total area of perfused vasculature in a given region of measurement. In addition to the length, it also involves the caliber of the vessels. Perfusion density is a relation of the number of pixels with perfused vasculature to the total number of pixels in the region; thus, it is expressed in %. Central vessel and perfusion densities (VDC and VPC, respectively) represent the parameters in the central circle of 1 mm in diameter. On the other hand, full vessel and perfusion densities (VDF and VPF, respectively) refer to the central area of 3 mm in diameter.
The study was approved by a local bioethics committee (Komisja Bioetyczna at OIL at OIL in Gdańsk; approval no.: KB-18/20 dated 7 July 2020).

Statistical Analysis
Statistical analysis was performed using the IBM SPSS Statistics 27 software. (IBM, Armonk, NY, USA) The effects of cataract surgery on retinal parameters were assessed by the repeated measures analysis of variance (ANOVA) for the four time points with the inclusion of the EPT covariant (ANCOVA). The analysis included the following parameters: BCVA, CRT, CRTA, and CV. Each of these parameters was assessed with the same ANCOVA model, and an additional ANOVA Friedman rank test was performed to confirm the differences in cases with high measurement errors. Multipair comparisons were performed with the Bonferroni correction. Statistical significance was reported for a p value of 0.05 or lower. OCTA measurements VDC, VDF, VPC, and VPF were evaluated with linear analysis of variance (linear ANCOVA) due to the relatively small sample.

Results
The descriptive statistics of the study group with the Shapiro-Wilk test results for the normality of distribution (p values) are presented in Table 1.

Best Corrected Visual Acuity
There was a significant improvement in BCVA after cataract surgery. The improvement was noted at each follow-up visit (p < 0.001). An increasing tendency was maintained at 2 weeks and 3 months, and then the BCVA values remained stable at 1 year after the surgery. There was no difference in BCVA at 3 months and 1 year (p = 1.000). The EPT did not significantly affect changes in BCVA during follow-up (ANCOVA F(1.42) = 0.21; p = 0.649; η 2 = 0.01).

Central Retinal Thickness
The ANCOVA assessing changes in CRT after the surgery revealed a result of borderline significance (F(3.126) = 2.55; p = 0.086; η 2 = 0.06). However, the additional ANOVA Friedmann rank test confirmed significant changes (F = 24.49; df = 3; p < 0.001). The analysis of simple effects with the Bonferroni correction showed that baseline CRT was significantly lower compared with values at 2 weeks (p = 0.028), 3 months (p < 0.001), and 1 year (p < 0.001). There were no significant differences between CRT values at different follow-up time points after the surgery. Variations in CRT after the surgery are presented in Figure 2.

Central Retinal Thickness
The ANCOVA assessing changes in CRT after the surgery revealed a result of borderline significance (F(3.126) = 2.55; p = 0.086; η 2 = 0.06). However, the additional ANOVA Friedmann rank test confirmed significant changes (F = 24.49; df = 3; p < 0.001). The analysis of simple effects with the Bonferroni correction showed that baseline CRT was significantly lower compared with values at 2 weeks (p = 0.028), 3 months (p < 0.001), and 1 year (p < 0.001). There were no significant differences between CRT values at different followup time points after the surgery. Variations in CRT after the surgery are presented in Fig

Macular Volume
The ANCOVA showed a significant increase in CV only at 3 months after the surgery (p = 0.003). However, the additional Friedmann rank test revealed significant changes in CV values at each time point (F = 24.68; df = 3; p < 0.001), with a p value of 0.008 at 2 weeks, < 0.001 at 3 months, and 0.028 at 1 year. The ANCOVA did not show a relationship between EPT and variations in CV (F(1.42) = 0.03; p = 0.863; η 2 < 0.01). Variations in CV after the surgery are presented in Figure 3.

Macular Volume
The ANCOVA showed a significant increase in CV only at 3 months after the surgery (p = 0.003). However, the additional Friedmann rank test revealed significant changes in CV values at each time point (F = 24.68; df = 3; p < 0.001), with a p value of 0.008 at 2 weeks, <0.001 at 3 months, and 0.028 at 1 year. The ANCOVA did not show a relationship between EPT and variations in CV (F(1.42) = 0.03; p = 0.863; η 2 < 0.01). Variations in CV after the surgery are presented in Figure 3.

Average Central Retinal Thickness
The ANCOVA did not reveal significant variations in CRTA after the surgery, but significance was revealed by the Friedmann rank test (F = 20.48; df = 3; p < 0.001). CRTA increased at 2 weeks (p = 0.012), 3 months (p < 0.001), and 1 year (p = 0.020) after the surgery, although the effect was weak. The EPT had no impact on variations in CRTA (F(1.42)

Average Central Retinal Thickness
The ANCOVA did not reveal significant variations in CRTA after the surgery, but significance was revealed by the Friedmann rank test (F = 20.48; df = 3; p < 0.001). CRTA increased at 2 weeks (p = 0.012), 3 months (p < 0.001), and 1 year (p = 0.020) after the surgery, although the effect was weak. The EPT had no impact on variations in CRTA (F(1.42) < 0.01; p = 0.999; η 2 < 0.01). Variations in CRTA after the surgery are presented in Figure 4.

Average Central Retinal Thickness
The ANCOVA did not reveal significant variations in CRTA after the surgery, but significance was revealed by the Friedmann rank test (F = 20.48; df = 3; p < 0.001). CRTA increased at 2 weeks (p = 0.012), 3 months (p < 0.001), and 1 year (p = 0.020) after the surgery, although the effect was weak. The EPT had no impact on variations in CRTA (F(1.42) < 0.01; p = 0.999; η 2 < 0.01). Variations in CRTA after the surgery are presented in Figure 4. .

Vessel Density Central
The linear ANCOVA test did not reveal significance of variations in VDC after the surgery (F(3,60) = 0.26; p = 0.857; η 2 = 0.04), nor was the significant effect of the covariant of EPT time proven (p = 0.243). Variations in VDC after the surgery are presented in Figure 5.

Vessel Density Central
The linear ANCOVA test did not reveal significance of variations in VDC after the surgery (F(3,60) = 0.26; p = 0.857; η 2 = 0.04), nor was the significant effect of the covariant of EPT time proven (p = 0.243). Variations in VDC after the surgery are presented in Figure  5.

Vessel Density Full
Analysis of variations in the VDF parameter after the surgery with the covariant of EPT did not show statistical significance (F(3,60) = 2,21; p = 0.096; η 2 = 0.20). However there was a weak relationship between higher EPT values and an increase in VDF (F(3,60) = 4.20; p = 0.045; η 2 = 0.36; r = 0.22). Variations in VDF after the surgery are presented in Figure 6.

Vessel Density Full
Analysis of variations in the VDF parameter after the surgery with the covariant of EPT did not show statistical significance (F(3,60) = 2,21; p = 0.096; η 2 = 0.20). However there was a weak relationship between higher EPT values and an increase in VDF (F(3,60) = 4.20; p = 0.045; η 2 = 0.36; r = 0.22). Variations in VDF after the surgery are presented in Figure 6.

Vessel Perfusion Central
The

Vessel Perfusion Central
The linear ANCOVA analysis did not show significant variations in VPC after the operation (F(3,60)

Vessel Perfusion Full
The linear ANCOVA test revealed significant variations in VPF after the surgery (F(3,60) = 2.98; p = 0.038; η 2 = 0.21). Nevertheless, there was no covariant effect of EPT (p = 0.106). The analysis of simple effects with Bonferroni correction showed a significant increase in VPF at each follow-up time point (p < 0.05). No significant differences among post-surgical VPF measurements were noted (p = 1.0). Variations in VPF after the surgery are presented in Figure 8.

Vessel Perfusion Full
The linear ANCOVA test revealed significant variations in VPF after the surgery (F(3,60) = 2.98; p = 0.038; η 2 = 0.21). Nevertheless, there was no covariant effect of EPT (p = 0.106). The analysis of simple effects with Bonferroni correction showed a significant increase in VPF at each follow-up time point (p < 0.05). No significant differences among post-surgical VPF measurements were noted (p = 1.0). Variations in VPF after the surgery are presented in Figure 8.

Vessel Perfusion Full
The linear ANCOVA test revealed significant variations in VPF after the surgery (F(3,60) = 2.98; p = 0.038; η 2 = 0.21). Nevertheless, there was no covariant effect of EPT (p = 0.106). The analysis of simple effects with Bonferroni correction showed a significant increase in VPF at each follow-up time point (p < 0.05). No significant differences among post-surgical VPF measurements were noted (p = 1.0). Variations in VPF after the surgery are presented in Figure 8.

Summary of the Results
In our study group, BCVA improved significantly after the surgery. BCVA achieved normal values in almost all patients (mean 0.02 ± 0.06 logMAR) at 3 months after the surgery, and the effect was maintained at 12 months (mean 0.01 ± 0.03 logMAR).
The statistical analysis confirmed significant postoperative changes in retinal thickness. CRT was increased at each follow-up time point compared with baseline. CV showed a significant increase at 3 months in the ANOVA and an increase at all three follow-up time points in the Friedmann rank test. CRTA also increased significantly at all three time points after the surgery, although the effect had moderate power. The increase in retinal thickness and volume had a tendency to rise in the first few months after the surgery and then to drop in the subsequent months. Nevertheless, the values at 12 months were still higher compared with baseline.
There were no significant changes in variations in VDC, VDF, and VDC after the surgery. On the other hand, vascular perfusion in the full retinal area (3 mm 2 ) improved after treatment. The increase was noted already at 2 weeks after the surgery and was maintained without fluctuations at subsequent time points. The question of why the improvement in vascular perfusion after the surgery was noted at the larger central retinal area but not in the smaller central part of 1 mm 2 has to be addressed. It should be remembered that retinal vascularity in the central 1 mm 2 naturally presents lower vessel density due to the presence of the foveal avascular zone; thus, changes in perfusion or vascular network density are more difficult detect [28]. This is why a larger sample is probably needed to confirm the perfusion improvement in this area.

SD-OCT and OCTA Measurements
When interpreting SD-OCT and OCTA measurements, the technical limitations of the procedure have to be considered. In our study, the retinal thickness and vascular perfusion parameters at baseline were lower than those after the surgery. Obviously, the presence of cataract affects the quality of the scans. Therefore, it is necessary to address the possible bias resulting from the poorer quality of presurgical scans. The ANOVA showed variations in scan quality, but the effect had moderate statistical power.
According to the manufacturer's recommendations, the threshold for signal strength accepted in our study (6/10) is also acceptable for the analysis of SD-OCT macular scans [29,30]. Previous studies showed that poorer quality of OCT scans may affect the outcome of the measurements, even if the signal strength or quality index stays above the threshold indicated by the device provider. However, these reports usually refer to optic nerve evaluation [31,32]. There is evidence showing that, in the case of macular cube scans, the lower quality affects the measurements to a lower degree [33]. Therefore, considering these reports, as well as the high mean and median values of signal strength in our study, we believe that the bias of low-quality scans for SD-OCT measurements can be treated as insignificant.
The issue of image quality is especially important for OCTA evaluation, as OCTA is burdened with the problem of artifacts. Most reports confirmed the impact of poorerquality OCTA scans on the measurement of vascular parameters, especially for repeated measurements [34][35][36]. Some authors directly recommend OCTA scan quality as high as 9/10 for comparative analysis [37]. In order to avoid bias of poor scan quality, we adopted the same high threshold for OCTA analysis.

EPT Time and Results after the Surgery
The ANCOVA revealed the significant effect of the EPT only on variations in VDF and not on any other retinal parameter. Additionally, this positive EPT effect was weak in power, so, taking into account the small sample of patients in this analysis, it needs to be confirmed in a larger study. In our study, cataract was not graded according to its density before treatment. Therefore, considering the acceptance of only high-quality OCT scans for the study, it can be assumed that hard and dense cataracts were excluded.

Other Studies
Our results should be placed within the context of other studies. Notably, our study has the strength of a long follow-up, which was 12 months. Studies with a follow-up of 6 months and longer reported similar results. A transient increase in CRT was typically observed after cataract surgery, with a peak value at around 2-3 months, followed by a drop to lower or baseline values [13,[15][16][17][18]20]. In our study, CRT at the end of follow-up was still higher than before cataract phacoemulsification.
There are scarce data on changes in OCTA parameters after cataract extraction (vascular density and vascular perfusion), with only a few studies available [20][21][22][23][24][25]. In a large study of 55 eyes, Krizanovic et al. [20] reported a stable improvement in vessel length and area, which was most prominent in the superficial vascular complex, noted from week 1 through month 3 after cataract extraction. Similarly, Zhao et al. [22] showed a significant increase in vascular density at 3 months after treatment in a study of 32 eyes. Zhou et al. [25] revealed a significant improvement in macular and peripapillary vascular parameters in a 1-month study of 51 eyes. The increase was noted after 1 week from surgery. Feng et al. [24] described an increase in superficial capillary plexus parameters in diabetic patients (32 eyes) 3 months after the surgery, but not in patients without diabetes. On the other hand, a small study of nine eyes by Pilotto et al. [21] showed only a transient increase in OCT and OCTA parameters, with a return to baseline at day 90 after the surgery. Interestingly, Li et al. [23] reported an improvement in vascular parameters 3 months after the surgery in patients with low myopia (31 eyes) and a worsening in patients with high myopia (24 eyes).
The impact of phaco energy during cataract surgery on retinal thickness parameters has not been widely investigated. Kurt et al. [12] reported a positive correlation between EPT and an increase in ganglion cell layer (GCL) and inner nuclear layer (INL) thickness. Similarly, Gołębiowska et al. [16] described higher values of retinal thickness and volume with higher phaco power used during the surgery, but not with longer phaco time. Other studies did not confirm the correlation between intraoperative factors (phaco energy, phaco duration) and retinal thickening after the surgery [8,17].
Thus far, the relationship between changes in vascular retinal parameters and the amount of phaco energy used during the procedure has been assessed only in the shortterm study by Zhou et al. [25]. The authors reported an increase in vessel density and perfusion density with larger values of cumulative dissipated phaco energy used during the procedure. However, they did not provide an explanation of the cause of this phenomenon. In our study, we found only a weak correlation between a higher EPT and VDF, and not with any other vascular retinal parameters. As our sample for the OCTA analysis was relatively small, this finding has to be interpreted with caution and confirmed in a larger trial.
An additional issue that should be taken into consideration when analyzing the impact of cataract surgery on postoperative retinal parameters is the type of pump implemented in the phaco machine. To our knowledge, this issue has not been addressed in the medical literature so far; therefore, only theoretical considerations can be presented. The two main types of pumps used in devices for cataract surgery are the Venturi pump and the peristaltic pump. A few available comparative studies showed that the Venturi pump requires smaller energy levels and a shorter time to complete cataract surgery than the peristaltic pump [38,39]. The Venturi type is also associated with a lower incidence of thermal corneal burns related to excessive thermal energy levels during the surgery [39]. The Stellaris phaco machine used in our study utilizes the Venturi pump; thus, we employed the technical option that was less likely to produce excessive energy levels during the surgery. If the higher phaco energy used during cataract surgery were to be related to postoperative retinal vascular changes, then it would be more likely to occur with the use of the peristaltic pump. However, this issue requires further comparative studies.

The Pathophysiological Concepts
The increase in retinal thickness and improvement in vascular parameters after cataract surgery could be attributed to a few processes. First, it could be due to the inflammatory reaction triggered by the surgery itself and hampered routinely by the application of topical anti-inflammatory drugs, as widely described in the literature [40,41]. Typically, inflammation after cataract surgery reaches its peak within a few days postoperatively and then declines over 2-3 weeks [42]. Our study showed that variations in retinal parameters are long-standing and, as such, they are quite unlikely to result from an inflammatory process induced by surgery. Moreover, none of the patients included in the study demonstrated any signs of chronic inflammation during the follow-up, including cystoid macular edema. The second postulated mechanism is associated with a decrease in intraocular pressure noted after cataract surgery, which could improve retinal perfusion and increase OCTA parameters [43][44][45]. While we did not address this issue in our study, other authors who reported postsurgical improvements in retinal vessels did not identify a relationship between a reduction in intraocular pressure and an increase in ocular perfusion pressure [20,25]. The third mechanism underlying the vascular improvement after cataract surgery is functional hyperemia, which is induced by an increase in light exposure leading to enhanced retinal metabolism [5,46]. Increased retinal metabolism involves the consumption of large oxygen and glucose amounts, which triggers the production of vasoactive mediators leading to vasodilation and hyperemia [6]. Logically, this effect should be reflected in an increase in OCTA parameters. According to our results, the improvement in retinal perfusion after cataract surgery is stable and long-standing, which indicates the third mechanism as the most plausible.
Nevertheless, it is important to note that despite an improvement in vascular perfusion, the higher exposure to light can also trigger angiogenesis, especially in susceptible eyes with signs of age-related macular degeneration. Further research on specific subgroups of patients is needed to establish whether the benefits of enhanced light exposure after cataract extraction outweigh the potential risks.

Study Limitations
The study included a relatively small number of cases, especially in the OCTA analysis, and its results have to be confirmed by large clinical trials. However, most current OCTAbased studies on this topic include a relatively small number of subjects. A dditionally, as dense cataract significantly reduces the quality of OCTA scans, the issue of applying the results to these cases remains to be clarified. Finally, the decrease in intraocular pressure after the surgery was not assessed as a potential factor influencing the improvement in vascular retinal parameters.

Conclusions
Uncomplicated cataract surgery in patients without retinal diseases results in an increase in retinal thickness and volume, with a peak value at around 3 months after the surgery and a subsequent decline. Vascular perfusion in the central 3 mm 2 of the macular area assessed by OCTA shows a stable improvement after the surgery. The EPT has no apparent impact on the variations in most of the retinal parameters after cataract surgery.

Institutional Review Board Statement:
The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by a local bioethics committee (Komisja Bioetyczna at OIL in Gdańsk, approval no. KB-18/20, dated 7 July 2020).

Informed Consent Statement:
Informed consent was obtained from all the subjects involved in the study.