Introduction

Vitreoretinal pathologies such as macular holes or epiretinal membranes are often associated with reduced reading ability and metamorphopsia.  Additionally, a concurrent senile cataract can also negatively affect visual acuity and the ability to visualize retinal pathology during examination.   In order to achieve rapid visual improvement for these patients and increase the surgical field of view, a combined approach incorporating cataract removal, followed by pars plana vitrectomy (PPV) is becoming the preferred approach for vitreoretinal surgeons. Combining these procedures minimizes the risk of postoperative PPV cataract development and provides the opportunity to completely fill the vitreous cavity with retinal tamponade. In concurrent cataract and retinal procedures, it is extremely important to confirm stable intraocular lens (IOL) position, prior to initiating PPV.   In eyes with a compromised red reflex, the creation of a well-centered continuous curvilinear capsulorhexis can be difficult [1,2]. Any deviation in size or centration of the IOL can cause it to tilt during further surgical manipulation and require repositioning following a fluid-air or gas exchange, which can be challenging. Further manipulation of the IOL postoperatively can lead to complications such as iris-capture and negatively affect visual performance.

Recent publications have provided evidence that cataract surgery combined with Femtosecond laser-assistance (FLACS) is a safe and reliable new surgical approach [3-5]. Intra- and early postoperative results comparing standard phacoemulsification with FLACS in combination with PPV indicated that the laser approach provides a stable IOL position with fewer complications during the first week of postoperative follow-up [6]. Until now, no long term data has been available comparing standard phacoemulsification with FLACS in combination with PPV.  The current study has assessed postoperative corrected distance visual acuity (CDVA), postoperative anterior chamber stability and evaluated potential impact of FLACS combined with PPV on endothelial cell loss, corneal thickening and macular thickness over 12 months of follow-up, compared with standard optimized phacoemulsification and PPV.

Methods Study design and oversight

In this prospective, randomized study, all patients underwent combined cataract surgery and PPV to treat surgery retinal disease, performed by a single surgeon at one clinical site. The study was reviewed and approved by the local Institutional Review Board (reference S-606/2016) and study conduct complied with the Declaration of Helsinki and followed International Conference on Harmonization (ICH) Guidelines for Good Clinical Practice. All participating patients provided written informed consent and understood that study participation was voluntary.

Study treatment

The details of the surgical techniques, diagnostic procedures, inclusion criteria and intraoperative and early postoperative complications have been described previously [6]. Eyes meeting the inclusion criteria were randomized (1:1) to undergo FLACS or conventional phacoemulsification cataract surgery prior to PPV. Randomization was performed using a computer-generated sequence (provided from the institutional IT-service).

Following randomization, those eyes in the conventional phacoemulsification group (n = 66) underwent cataract removal with Stellaris Elite (Bausch & Lomb, USA). For those randomized to FLACS (n = 65 eyes) the femtosecond laser procedure was performed with Catalys Precision Laser System (Johnson & Johnson, USA) to create the capsulotomy and perform lens fragmentation [6]. A monofocal, single-piece IOL (TECNIS PCB00, Johnson & Johnson) was implanted in all eyes prior to proceeding to PPV. A 23-gauge PPV was performed in all eyes, using the Stellaris Elite platform.  According to the underlying retinal pathology, membrane peeling, removal of the ILM, fluidair exchange and gas endotamponade was performed.

To ensure comparability between treatment groups, the volume of irrigation fluid used during lens removal for each group was recorded.  Following surgery, the unused fluid volume was measured and subtracted from the original 500ml balanced salt solution (BSS) bottle volume. The fluid volume used for priming of the phacoemulsification machine prior to surgery is constant and had not to be calculated separately. Additionally, the surgical time for the phacoemulsification procedure without PPV was documented. For this study cohort, effective phacoemulsification time (EPT), additional surgical steps required to regain stable IOLposition, intra- and early postoperative overlap of capsulorhexis/ capsulotomy and IOL optics have been previously published [6].

Study assessments

The primary outcome of the study was the change in corrected distance visual acuity from baseline (pre-operatively) to 12 months postoperatively.  Secondary outcomes included changes in corneal pachymetry and corneal endothelial cell counts, anterior chamber and macular parameters, as well as anterior and posterior segment intra- and postoperative complications. Study assessments were performed preoperatively and postoperatively by one of two masked physicians at Day 10, Month 1, Month 3, Month 6, and Month 12 follow-up visits.  Best-corrected distance visual acuity was measured using Early Treatment of Diabetic Retinopathy Study (ETDRS) charts.  Heidelberg Spectralis optical coherence tomography (OCT) (1.8.6.0, Heidelberg Engineering, Germany) was used to assess macular parameters, including mean retinal thickness,  central macular thickness  (6.0 mm macular zone) as well as central foveal thickness (1.0 mm diameter central circle), per parameters based on ETDRS. Endothelial cell count (ECC) and corneal thickness were measured preoperatively, and 1 day and 3-4 days postoperatively (only pachymetry), 7-10 days, 5060 days, 90-100 days and one year postoperatively by one of two masked physicians. The Sea Eagle endothelial non-contact computer-assisted specular microscope (Rhine-Tec, Germany) was used to objectively assess ECCs.  Scheimpflug cross-sectioning anterior segment imaging (Pentacam HR, Oculus, Germany) was performed to provide corneal pachymetry data.  This allowed for calculation of absolute hexagonal cell values in addition to the percentage loss in ECC, using the previously published formula [7].

Statistical Analysis

The planned enrollment was based on the statistical power calculation of the previous study [7], which would provide >80% statistical power to detect a statistically significant difference between groups, by means of a 2-sample Wilcoxon test at a 5% significance level. Descriptive statistics were used to summarize all participant characteristics and pre-operative and post-operative assessment data, as well as changes over the study period. Results for each treatment group were depicted in box plots or presented in tables containing mean, standard deviation (SD), median, minimum and maximum values. T-tests were used to compare sample means and P-values less than 0.05 were considered to be statistically significant. Analyses was performed using SPSS Version 27.0 (SPSS Inc., USA).

Results

A total of 111 eyes were enrolled and randomized into the study. During the 12 month study, three eyes of three patients in both groups were lost to follow-up, did not complete their final followup visit, and were therefore excluded from the final analyses.  At the end of the study, 63 eyes of 62 patients received conventional phacoemulsification followed by PPV, and 62 eyes of 60 patients were treated by FLACS combined with PPV, completed their 12 month follow-up visit. 

Participant demographics, baseline characteristics and retinal pathologies are presented in Table 1.  No significant differences were observed between groups. At 12 month follow-up visit mean decimal DCVA was 0.70 (SD 0.27; median 0.63; 0.1 to 1.0) in the FLACS group in comparison to 0.65 (SD 0.27; median 0.80; 0.05 to 1.0) in the conventional phacoemulsification group. The changes in CDVA are presented in Figure 1. One eye of the conventional phacoemulsification, that underwent PPV for retinal detachment lost one line of CDVA, while all other eyes remained stable or gained lines in visual acuity at one year follow-up in comparison to assessed preoperative CDVA. Postoperatively, cystoid macular edema (CME) occurred in three eyes in the FLACS group (1 eye with a macula involved retinal detachment, 2 eyes with macular pucker) and five eyes in the conventional phacoemulsification group (2 eyes with retinal detachment, 1 eye with macular hole, 1 eye with macular pucker, 1 eye with vitreous haemorrhage associated with proliferative diabetic retinopathy).

Table 1: Demographics and Baseline Ocular Characteristics; *FLACS - femtosecond laser-assisted cataract surgery; ** Standard - conventional phacoemulsification.

Figure 1: Safety - preoperative vs one year postoperative CDVA in the FLACS group and in the standard group.

Two eyes of the FLACS group received additional panretinal laser coagulation (1 eye with retinal detachment, 1 eye with proliferative diabetic retinopathy) and one eye in the conventional phacoemulsification group with preoperative complete retinal traction detachment underwent a second surgery due to re-detachment of the retina four months after the initial surgery. Preoperative OCT could only be performed in eyes without retinal detachment of the macular area or dense vitreous hemorrhage. Although preoperative data was missing, these eyes were included in analysis of postoperative CME. Mean central retinal thickness and central foveal thickness values are illustrated in Figures 2 and 3.  Preoperatively mean central retinal thickness was 388 µm (SD 84) in the FLACS group and 414 µm (SD 107) in the conventional phacoemulsification group. One month postoperatively, the mean central retinal thickness in the FLACS group was 357 µm (SD 72), 351 µm (SD 65) at three months, 339 µm (SD 61) at six months and 327 µm (SD 60) at the 12 month followup. In the conventional phacoemulsification group mean central retinal thickness was 379 µm (SD 96) one month postoperatively, 369 µm (SD 80) at three months, 357 µm (SD 75) at six months and  343 µm (SD 74) at 12 months. Mean central retinal thickness and central foveal thickness were not significantly different between groups at each follow-up visit.

Figure 2: Boxplot of retinal mean center thickness from preoperative values over one year follow up time. The bottom and top of the box are the 25^th and 75^th percentiles, respectively, and the band near the center is the 50^th percentile (median). The bars outside the box indicate the maximum and minimum of all data. A minor outlier (denoted by a small circle) is an observation 1.5 x interquartile range outside the central box; The change is statistically significant not different over the whole postoperative period (p<0.001).

Figure 3: Boxplot of retinal central foveal thickness from preoperative values over one year follow up time. The bottom and top of the box are the 25^th and 75^th percentiles, respectively, and the band near the center is the 50^th percentile (median). The bars outside the box indicate the maximum and minimum of all data. A minor outlier (denoted by a small circle) is an observation 1.5 x interquartile range outside the central box; The change is not statistically significant different over the whole postoperative period (p<0.001).

The absolute endothelial cell counts did not demonstrate any statistically significant differences over time.   In the FLACS group the absolute endothelial cell values changed from 2431 cells/mm² (SD 312; range 1743 to 3152) preoperatively, to postoperatively, 2336 cells/mm² (SD 295) at one week, 2362 cells/mm² (SD 301) at three months, 2401 cells/mm² (SD 310) at six months and 2394 cells/mm² (SD 313; range 1702 to 3111) at 12 months. In the conventional phacoemulsification group, absolute endothelial cell values (Figure 4) changed from 2612 cells/mm² (SD 239; range 1980 to 2109) preoperatively, to 2361 cells/mm² (SD 256) at one week, 2358 cells/mm² (SD 251) at three months, 2349 cells/mm² (SD 245) at six months and 2366 cells/mm² (SD 260, range 1763 to 2801) at twelve months postoperatively.

Figure 4: Boxplot of absolute central endothelial cell count over twelve months follow-up time; The bottom and top of the box are the 25^th and 75^th percentiles, respectively, and the band near the center is the 50^th percentile (median). The bars outside the box indicate the maximum and minimum of all data. A minor outlier (denoted by a small circle) is an observation 1.5 x interquartile range outside the central box; The change is not statistically significant different over the whole postoperative period (p<0.001).

Relative change in corneal endothelial cell loss demonstrated significant differences between groups at all examination points and showed positive correlation with intraoperative EPT. Relative change in corneal endothelial cell loss from the preoperative evaluation (Figure 5) to one week postoperatively was 3.8 % (SD 3.2), 2.8 % (SD 2.7) after three months, 1.2% (SD 1.7) after six months and 1.5% (SD 3.5) at twelve months postoperatively in the FLACS group, compared to the conventional phacoemulsification group, which showed 9.6 % (SD 4.7) relative change in corneal endothelial cells loss at one week, 9.7 % (SD 4.7) after three months, 10.0 % (SD 4.8) after six months and 9.5 % (SD 5.2) at twelve months postoperatively. Corneal thickness was only significant different between groups at Day 1, Day 3, and Day 10, but not at any later time points (Figure 6).

Figure 5: Boxplot of corneal endothelial cell loss (%) from preoperative values over twelve months follow-up time; The bottom and top of the box are the 25^th and 75^th percentiles, respectively, and the band near the center is the 50^th percentile (median). The bars outside the box indicate the maximum and minimum of all data; A minor outlier (denoted by a small circle) is an observation 1.5 x interquartile range outside the central box; A major outlier (denoted by a *) is an observation 3.0 x interquartile range outside the central box; The change is not statistically significant different over the whole postoperative period (p<0.001).