One-year results and safety of femtosecond arcuate incisions for high post-keratoplasty astigmatism treatment: time to rethink the procedure?


 Background/Aims High astigmatism after keratoplasty is the main limiting factor for visual rehabilitation. Among its treatment options the arcuate keratotomy (AK) with the femtosecond laser stands out for its perfection and versatility of the incision making. We designed a study to evaluate the results and safety of femtosecond arcuate keratotomy (femto-AK) in the treatment of high astigmatism after keratoplasty.Methods A prospective interventional cohort study. We enrolled 17 eyes with high degree of irregular astigmatism, scheduled for Femto-AK. IntraLase FS was used to perform arcuate incisions 1.00 mm inside the graft. Patients’ uncorrected visual acuity (UCVA), best spectacle corrected visual acuity (BDCVA), and astigmatic change were recorded and followed up to one year after surgery.Results Despite overall improvement in corneal cylinder and visual acuity in one-year follow-up, there was a 52.9% complication rate. Cylinder overcorrection occurred in 6 cases (35.3%) that was managed with compressive sutures; microperforation occurred in two cases (11.7%); and endothelial rejection in one case (5.9%).Conclusion The predictability of astigmatism correction was variable in reducing postkeratoplasty astigmatism and high complication rate was found in our case series. New nomograms and improvements in technique need to be developed in order to reduce the incidence of overcorrections.


Introduction
Currently, success rates of keratoplasty, de ned as transparency of the cornea one year postoperative, are above 90% [1].
However, a transparent cornea does not always mean satisfactory visual outcome, being postoperative astigmatism the main responsible factor for an unsatisfying visual rehabilitation [2,3]. The high astigmatism has a negative impact on quality of life of cornea transplanted patients [4] and may be related to various factors as quality of trepanning, disparities between donor and recipient cornea, irregularities in the incision con guration, recipient underlying disease, suture technique, and postoperative control of astigmatism [5].
In the rst year after corneal transplantation, the most widely used method to control the astigmatism is the selective removal of corneal sutures. Despite this, only 27% to 34 % of patients have astigmatism less than or equal to 3.0D in two years [6]. When high astigmatism persists after all sutures are removed, and when the patient does not tolerate hard contact lenses, surgical methods of postoperative astigmatism reduction can be used, including wedge resection [7], compression sutures, intrastromal ring segments, refractive surgery with excimer laser, [8] and arcuate keratotomies (AK) [9].
For mixed or simple myopic astigmatism, AK may represent an interesting option [10]. Manual AK, including freehand techniques [9] or mechanical techniques using devices such as Hanna arcitome [11], is an established method of addressing high degree of post-keratoplasty astigmatism that cannot be managed by spectacle lenses or hard contact lenses. The main principle of AK is based on attening the steep corneal meridian by placing 1 or 2 incision at the steep axis. This attens the steep meridian with a reciprocal steepening of the at meridian 90 degrees away, known as a "coupling effect" [9].
In 2006, femtosecond laser was rst used to create PKP incisions (top-hat wound con guration) [12] and to perform femtosecond laser-assisted astigmatic keratotomy [13]. Femtosecond laser arcuate incisions are more precise in arc depth, diameter, length, and curvature [14]. This technique seems to be better than manual keratotomy in terms of safety and fewer higher-order aberrations [15,16,17,18].
This prospective interventional study describes our results in safety and e cacy of femtosecond laser-assisted AK to treat high astigmatism after keratoplasty.

Patients
We have included 17 consecutive patients (8 males, 9 females) from May 2014 to July 2015. Institutional ethics committee approval and informed consent for all participants were obtained for this study. Eyes with at least 5.0 of postkeratoplasty astigmatism were included. The time between keratoplasty and AK varied from 1 to 25 years. The keratotomies were performed after at least 2 years of all suture removal. Sixteen patients were submitted to penetrating keratoplasty while one patient was submitted to deep anterior lamellar keratoplasty.
Inclusion criteria included a clear cornea graft and a corneal topography with a regular, symmetric or asymmetric, astigmatism.
Outcome measures included BSCVA, manifest refraction and Pentacam® tomographic measures. Patients were followed postoperatively at 1 day; 1 week; and 1, 3, 6 and 12 months and so each year or less if necessary. Postoperatively, patients were treated with a topical combination of 0.3% gati oxacine and 1% prednisolone acetate drops (Zypred, Allergan®), four times a day for 1 week and then were placed on a maintenance dose of a topical steroid according to their individual requirements, as well as lubricant eyedrops.
The laser's limbal suction ring was applied and cone positioned. Applanation was considered adequate if the uid meniscus was at least beyond the graft-host junction (hard docking).
All surgeries were performed using a single or a paired asymmetric incision centred on the topographic location of the steep meridian, according to Hurmeric and Yoo nomogram [24]. The optical zone ranged from 6.00 to 6.75 mm in diameter and the incision depth was 90% of Scheimp ug pachymetric measurement in 8 cases and 85% in 9 cases (incision depth ranged from 410 to 567um). Lateral cut angles ranged from 60 to 90 degrees and anterior side-cut angle was set at 90 degrees in all cases.
The change in astigmatism was assessed by absolute cylinder variation in refraction and by Scheimp ug tomographic measurements.
Statistical Analysis SPSS software (version 18.0, SPSS, Inc.) were used for statistical analysis. A paired t test was used to compare continuous variables. Data were expressed as means ± SD, and P value of less than 0.05 was considered statistically signi cant.

Results
Seventeen eyes of 17 patients (8 men, 9 women) were included in the study. The mean age of patients was 40 years.
Keratoconus was indication for keratoplasty in 15 cases (88%). The mean interval from keratoplasty to the AK procedure was 8.2 years.  xpressed as mean ± SD . s are referred to preoperative measurements.
Mo, months; P, p value; D, diopters, BSCVA, best spectacle-corrected visual acuity; SD, standard deviation Overcorrection occurred in 6 cases. In one, there was an initial satisfactory result, but 4 months postoperatively the arcuate incisions opened, and the topographic astigmatism was inverted by 10D (Table 3, patient no. 9). The patient was submitted to compressive sutures with good outcome up to now. In another case stromal oedema has occurred adjacent to one of the keratotomies, inducing a 11D of topographic astigmatism, that was managed with topical steroids and compressive sutures, with improvement over time. In the other four cases, the opened incisions resulted in an inversion of astigmatism in great magnitude and were managed by compressive sutures with good results. Microperforation occurred in two cases during keratectomy; It was managed with therapeutic contact lens without other complications.

Discussion
Arcuate incisions are considered an e cient method in the treatment of high astigmatism after keratoplasty. From its conception by Snellen [19] in 1869 to the beginning of its clinical application in the 1970s by Troutman [25], Krachmer [26] and Lindstrom [27], among others, the procedure has shown good results for a di cult-to-treat problem until then. Nevertheless, undesirable complications such as microperforations and overcorrections may occur in some cases.
The advent of femtosecond laser has created an important expectation in the improvement of the technique, due to its greater precision and perfection in the incision making. Since then, some results are being published about the results of the arcuate incisions made with the new technology. Irit Bahar et al. [15] compared the results between manual and femtosecond laser techniques and found no statistically signi cant difference regarding best-corrected visual acuity, uncorrected visual acuity and mean absolute cylinder. Moreover, there was no difference in overcorrection rate, which occurred in 25% in the cases of the femto-AK group.
In our case series, unexpectedly, we also had a high frequency (35%) overcorrection, 33% of these occurring 6 months after the procedure. Maybe because the incisions are so perfect that the healing process is weaker than when manually performed. One option to try to avoid this is a technique proposed by Clearly et al. [21] by making a 135º beveled incision rather than a perpendicular one. Likewise, the author still recommends more studies with a larger number of patients to adopt the technique.
In fact, overall, the results of the AK procedure, whether mechanized or Femto-AK, show some success in correcting the refractive error, but with varying unpredictable results. In order to have more consistent results, Nubile et al. [16] limited incisions to 80 arc length and had no cases of overcorrection in a 10-case series, indicating that perhaps revision of the traditional nomogram is advised to regard the use of high arc length incisions. St Clair et al. [20] proposed modifying Hanna's nomogram based on regression analysis of a large group of patients from multiple centres. However, this theoretical model still needs to be applied and its results evaluated in a prospective study.
Another possibility to prevent overcorrection overtime due to incision opening is to perform intrastromal femto-AK, which seem to be less effective, but safer. Viswanathan et al. reported signi cant reduction in high postkeratoplasty astigmatism with femtosecond intrastromal arcuate keratotomies [27].
If, on the one hand, the femtosecond laser for arcuate incisions did not show clinically superior results compared to the manual technique, in the other hand, its superiority with regard to the accuracy and safety of the instrastromal corneal ring (ICRS) implantation is undoubted. In this sense, the use of ICRS proved to be an excellent technique for the correction of high astigmatism after keratoplasty. Coscarelli et al. [22] showed effective corneal cylinder reduction and no cases of visual loss in a series of 59 eyes. In the same way, Arriola-Villalobos et al. [23] published very similar results showing the consistency of the technique.
In conclusion, we have found a high overcorrection rate over time with femtosecond laser arcuate incisions in the treatment of high astigmatism after keratoplasty, many of them needed to be sutured. From our perspective, better nomograms for this technique need to be studied.

Declarations
Contributors: All authors made substantial contributions to the research. SK was the principal investigator. All authors contributed to the data collection, design, analysis and interpretation of the data, and to the revision of the manuscript.
Funding: None