Changes in Ganglion Cell Complex and Peripapillary Retinal Nerve Fiber Layer after Femtosecond Laser-Assisted Cataract Surgery Compared to Manual Phacoemulsification in Patients Receiving a Trifocal Intraocular Lens

Introduction During femtosecond laser-assisted cataract surgery (FLACS), there is a significant increase in intraocular pressure, which might lead to ganglion cell damage. We aimed to determine whether there were differences in the changes produced in the ganglion cell complex (GCC) and peripapillary retinal nerve fiber layer (pRNFL) thickness, as evaluated with optical coherence tomography (OCT), between phacoemulsification and FLACS, after implantation of a trifocal intraocular lens (IOL). Methods Patients with no coexistent pathologies undergoing cataract surgery with implantation of a PanOptix IOL were explored with the Cirrus-OCT before and three months after surgery. GCC values were obtained from the built-in software. The differences between pre- and postoperative GCC and pRNFL thicknesses after phacoemulsification were compared to differences after FLACS. Results A total of 171 eyes were included, 74 undergoing FLACS and 97 phacoemulsification. For both groups, there was a statistically significant increase in GCC values after cataract surgery, except for the inferior and inferonasal sectors. There were no statistically significant differences between FLACS and phacoemulsification. Mean change in average GCC and minimum GCC were 1.08 ± 1.40 µm (range −1 to +6 µm) and 1.69 ± 2.54 µm (range −3 to +11 µm) after FLACS and 0.99 ± 1.67 µm (range −5 to +6 µm) and 2.02 ± 3.54 µm (−6 to +18 µm) after phacoemulsification. These values are similar to those previously reported after phacoemulsification with monofocal IOL implantation. No significant changes after surgery were detected for the pRNFL, with no differences between groups. Discussion. There were no differences in the changes produced by FLACS and phacoemulsification in either GCC or pRNFL values. Although mean change was small, the range of variation was wide. Therefore, it is necessary to establish a new baseline for GCC and pRNFL thicknesses after cataract surgery in order to monitor any subsequent changes.


Introduction
Optical coherence tomography (OCT) has become one of the mainstays in the evaluation of ophthalmic patients. In the management of glaucoma, detecting structural change over time in the peripapillary retinal nerve fiber layer (pRNFL) and in the ganglion cell complex (GCC) is useful for both the initial diagnosis and detecting progression. Many patients show structural changes in the absence of visual field deterioration, providing the opportunity to start or increase treatment before the onset of permanent visual damage [1,2]. Detecting change over time has advantages over comparing a single scan with a normative database, due to the fact that "normal" values vary widely between different subjects. us, it is possible for a patient to suffer significant neural loss before being deemed outside normal limits [3]. Several studies have reported that GCC and pRNFL thickness parameters significantly increase after cataract surgery [4][5][6][7] with a tendency to decrease with time [8], although without reaching preoperative levels [9,10].
Nowadays, femtosecond laser-assisted cataract surgery (FLACS) is expanding. It is well known that cataract surgery leads to a transient intraocular pressure (IOP) elevation in the early postoperative period [11]. In addition, the femtosecond procedure leads to a significant increase in IOP: experiments in ex-vivo porcine models have reported that the IOP increases during the treatment steps to up to 48.41 ± 6.80 mmHg with the Catalys platform [12], 52.00 ± 6.35 mmHg with the Femto LDV Z8 [13], and 77.01 ± 5.88 mmHg with the Victus platform [14]. An in vivo trial with IOP measurement with an I-Care tonometer found that IOP rose from a preoperative mean of 13.8 ± 0.4 mmHg to 24.2 ± 1.4 mmHg one minute after docking release in patients undergoing FLACS with a LenSx system [15]. In another in vivo trial, IOP rose to 42.1 ± 10.8 mm Hg during the suction phase with the Victus platform [16]. ere is some concern that these IOP spikes, in elderly patients with more vulnerable optic nerves, might lead to retinal nerve fiber damage. Animal models have found that even transient ocular hypertension may lead to early changes in the structure and function of various retinal ganglion cell types [17]. e purpose of this study was to evaluate the changes produced in GCC values after FLACS and compare them with changes produced after "classic" phacoemulsification, with the implantation of a trifocal intraocular lens (IOL), in order to determine whether IOP increase during the femtosecond procedure might lead to ganglion cell damage, reflected in OCT measurements.

Methods
All patients who underwent cataract surgery with implantation of a trifocal IOL AcrySof IQ PanOptix TM between January 2017 and June 2018 in our center were considered for inclusion. At the one-month postoperative visit, the nature of the study was explained and all those patients who agreed to participate and signed the informed consent were scheduled for a visit three months after surgery. e study was approved by our ethics committee and followed the tenets of the declaration of Helsinki.
Inclusion criteria were age over 18 years, preoperative OCT with signal strength ≥6, and uneventful cataract surgery with Panoptix IOL implantation. Exclusion criteria were presence of ocular pathologies, amblyopia, postoperative complications, and suboptimal segmentation of ganglion cell complex. Only one eye per subject was included in the study. e right eye was chosen unless preoperative signal strength was below required or there was incorrect segmentation, in which case the left eye was chosen.
Optical coherence tomography measurements were performed with the Cirrus HD OCT imaging system (Carl Zeiss Meditec, USA), before and three months after cataract surgery. For macular examination, the macular cube 512 × 128 acquisition protocol was used.
is protocol generates a cube through a 6 mm square grid of 128 B-scans, each composed of 512 A-scans. e built-in software identifies the outer boundaries of the RNFL and the inner plexiform layer. e difference between the RNFL and the inner plexiform layer outer boundary segmentations yields the combined thickness of the retinal ganglion cell/inner plexiform layer, which we will refer to as GCC. e Ganglion Cell Analysis algorithm provides GCC measurements in six wedge-shaped sectors centered on the fovea. It also gives information on the average and minimum GCC thickness for each eye and compares these figures with a normative database. Examples of a pre-and postoperative analysis are shown in Figures  e main corneal incision was fixed at 135°with a width of 2.3 mm and the sideport at 60°with a width of 1 mm. e capsulotomy had a diameter of 5 mm and a combined radial and cylinder pattern was employed for lens fragmentation.
Surgery was completed with the Centurion ® Vision system (Alcon Laboratories, USA). e Panoptix IOL was injected in the bag. For patients undergoing "classic" phacoemulsification, surgery was performed through a 2.2 mm clear corneal incision, with a Stop and Chop technique. Intracameral cefuroxime was used at the end of all procedures as prophylaxis for endophthalmitis.
Statistical analysis was performed with SPSS software (version 20.0, IBM SPSS, USA). e Kolmogorov-Smirnov test was used to evaluate the normal distribution of measured data. When parametric analysis was possible, Student's t test for paired data was employed to compare pre-and postoperative values for both groups and Student's t test for unpaired data was used for evaluating differences between the FLACS and phacoemulsification groups. e Bonferroni correction was applied due to the multiple comparisons performed, so that significance was set at p < 0.0015.

Results
A total of 171 eyes of 171 patients were included in the study. Of these, 74 eyes underwent FLACS (FLACS group) and 97 eyes "classic" phacoemulsification (Phaco group). ere was a higher percentage of women in the FLACS group (58/74 eyes, 78.4%) compared to the Phaco group (67/97, 69.1%), although the difference was not statistically significant, p � 0.223. Mean age was similar in both groups: 67.12 ± 7.80 years for FLACS versus 67.74 ± 7.93 for the Phaco group. Table 1 shows the preoperative characteristics of the eyes included in the study. Tables 2 and 3 show the pre-and postoperative values of the GCC for the Phaco and FLACS groups, respectively. For both groups, there was a statistically significant increase in all GCC values except for the inferior and inferonasal sectors. ere were no statistically significant differences in GCC changes between the FLACS and Phaco groups ( Table 4). As regards pRNFL thickness, there were no statistically significant changes with cataract surgery, neither with FLACS nor classic phacoemulsification (Tables 5 and 6).

Discussion
Optical coherence tomography examination of the optic nerve head and the macula is nowadays performed almost systematically as part of ophthalmic evaluations. Measurements of the pRNFL and the GCC thicknesses are used to diagnose glaucoma and to monitor the course of the disease. Several studies have reported that cataract surgery leads to changes in the GCC and pRNFL thickness values [4][5][6][7][8][9][10] and these changes must be taken into account for patient followup. However, to the best of our knowledge, no study has directly compared the effect of FLACS with "classic" phacoemulsification on GCC values. Since FLACS has been  Journal of Ophthalmology shown to be associated with a significant, although transient, increase in IOP, this comparison might detect differences in postoperative GCC and pRNFL thickness due to ganglion cell damage.
In our study, we found no differences in the changes produced by FLACS and "classic" phacoemulsification in GCC values. e mean increase in average GCC was 1.08 ± 1.40 µm for the FLACS group compared to 0.99 ± 1.67 µm for the Phaco group. is change in average GCC value is comparable to those reported in other studies evaluating changes after classic phacoemulsification with Cirrus-OCT, with increases ranging from 0.57 to 4.2 µm [4,7,8]. However, although the mean change across the whole study group is small, the range of variation we found was wide: for a given individual, average GCC can vary as much as 6 µm, minimum GCC as much as 18 µm, and the value for some sectors as much as 17 µm. erefore, it is imperative to establish a new baseline for GCC after cataract surgery in order to monitor any subsequent changes.
As regards changes in average pRNFL thickness, a mean increase of between 2.11 µm and 5.63 µm has been reported after classic phacoemulsification using a Cirrus-OCT   Journal of Ophthalmology [4,5,7,8]. In our study, we found that average pRNFL did not change significantly after cataract surgery, with no differences between the phaco and FLACS groups. is might be due to the larger number of patients included in our study, which might tend to compensate differences. Regarding the specific effect of FLACS, Reñones et al. reported a mean increase of 2.16 µm in average pRNFL as evaluated with Spectralis-OCT, 6 months after surgery [18]. eir study did not include a control group undergoing classic phacoemulsification. Although comparisons must be made with caution, since it has been reported that there are differences in RNFL measurements between devices [5], it seems that pRNFL change after FLACS is within the values reported for phacoemulsification. e reason for the changes in OCT measurements after cataract surgery is not completely clear. It seems that it may be due to a combination of several factors, including the inflammatory effect of the procedure, a decrease in the  Journal of Ophthalmology optical density of the lens, the change in refraction after surgery, and the optical properties of the IOL [5,6]. e implantation of trifocal IOLs is steadily increasing and it is necessary to determine how they might affect OCT measurements. García-Bello et al. [19] [19]. us, the greater difference found for trifocal lens might have been due to inaccurate preoperative measurements. e same group has published another study comparing the AT LISA tri839MP and the FineVision IOL (PhysIOL, Belgium), another trifocal IOL, with 24 patients per group [20]. In this study, average pRNFL thickness  increased in the Finevision group from 88.06 ± 13.65 µm to 92.04 ± 12.78 and in the AT LISA group from 82.39 ± 15.14 to 86.88 ± 10.95 µm. ese results are within the changes reported for monofocal IOLs. In our study, all patients received a Panoptix trifocal IOL; again, we believe the lack of a significant difference in mean average pRNFL thickness might be due to the larger number of patients included. However, as is the case for CCG measurements, although there is no significant difference in the mean average pRNFL, average pRNFL in a given patient might change as much as 10 µm and therefore it is necessary to acquire a new baseline after surgery for further follow-up.
e main limitation of our study was that patients were not randomized to FLACS or classic phacoemulsification. However, the groups had similar baseline characteristics. On the other hand, two of the strengths of our study are the large      number of patients included and the fact that only one eye per patient was analyzed.
In summary, we found that both FLACS and classic phacoemulsification lead to a small but statistically significant increase in most GCC measurements with Cirrus-OCT. No difference was found for average pRNFL measurements.
ere were no differences in the changes produced between FLACS and classic phacoemulsification. Although mean changes are low, individual variations mean that a new baseline should be acquired after surgery.
Data Availability e dataset analyzed during the current study is available from the corresponding author on reasonable request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.