Achieving emmetropia in eyes with full thickness corneal grafts is a target that has long been pursued. For decades, a successful keratoplasty was judged in terms of preservation of a clear surviving graft as a final outcome. However, the ultimate goal of vision restoration was often hampered by the frequent association with postoperative ametropia and/or anisometropia [
A wide scope of therapeutic modalities has been proposed and employed for the correction of such refractive errors. These ranged from the very conservative spectacle prescription to the final possibility of repeating the whole grafting procedure. However, none of these techniques has proven itself as a sole ideal solution for the management of post-PKP ametropia and many are associated with graft survival and, even, vision threatening complications [
Visual rehabilitation using spectacles represents a good choice but their use is limited by significant anisometropia especially with astigmatism more than 3 D or the presence of irregular astigmatism. Contact lenses (especially rigid gas permeable) provide another conservative management option but dry-eye syndrome and fitting-related inconvenience as well as patient’s age, dexterity, and lifestyle are major concerns that may affect contact lens tolerance. They may also induce peripheral corneal neovascularization which can result in graft rejection [
Initial surgical management options range from simple procedures like selective suture removal [
Intracorneal ring segment (ICR) implantation has recently been reported in post-PKP eyes but with significant undercorrection and low predictability compared to their results in keratoconic eyes. Immune reaction induction leading to graft rejection and ring migration and perforation into the anterior chamber are possible complications that can violate the future integrity of the grafted cornea [
Excimer laser photorefractive keratectomy (PRK) has also been proposed but was proven less predictable than when performed in previously nonoperated eyes. Limited astigmatic correction, irregular astigmatism induction, significant regression, corneal haze, and photoablation induced graft rejection are other important disadvantages of such a technique [
Toric intraocular lens implantation (phakic or pseudophakic) provides a wider range of correction but lens rotation, increased endothelial cell loss, corneal decompensation, graft rejection, endophthalmitis, and secondary glaucoma are important drawbacks of such procedures [
Small incision lenticule extraction (SMILE) is a novel, all-in-one, corneal laser refractive surgical procedure in which a lenticule of stromal tissue of planned thickness and diameter is isolated between two intracorneal planes created using a femtosecond laser platform. The lenticule is, then, manually removed from the cornea through a small incision to change the corneal curvature and exert its refractive effect. It was reported to be safe, predictable, and effective for treating myopia and myopic astigmatism in previously nonoperated eyes. It has the advantages of being flapless and less invasive than other intraocular procedures together with having the ability to tailor and center the whole procedure as required within a specific area of the cornea [
This study was performed with the aim of evaluation of visual and refractive outcomes after SMILE for treating post-PKP myopia and myopic astigmatism.
This interventional case-series included ten eyes of 10 patients with previous PKP and residual compound myopic astigmatism.
Inclusion criteria consisted of patients who had had an 8.25 mm donor button transplanted to 8.00 mm trephination of the recipient cornea with a duration of at least 18 months since the time of keratoplasty and a residual myopic refractive error of up to −10.0 diopters (D) of spherical equivalent with astigmatism up to 6.0 D at the spectacle plane. A smooth postkeratoplasty course with no attacks of graft rejection or suture complications was mandatory. Sutures were completely removed prior to performing the preoperative examination by at least three months during which patients were followed up monthly to ensure a stable refraction. Only patients whose topography and anterior and posterior elevation maps’ data within the acceptable range for laser vision correction and a thinnest graft location of 500
Patients with graft apposition abnormalities (override or underride), severe dry eye, ocular surface disease, abnormal topographies, thin grafts, elevated intraocular pressure (IOP), peripheral corneal neovascularization, nonsuture track related peripheral opacities, or central or paracentral opacities were excluded. Patients experiencing post-SMILE interface inflammation, cellular infiltration, or any other reported SMILE complication were planned to be excluded, as well. Also patients with other ocular or recorded eye-related systemic illnesses (e.g., diabetes mellitus) were not included in the study.
The study was conducted between 2011 and 2015. The study protocol was based on the tenets of the Declaration of Helsinki and was approved by the ethics committee of the Faculty of Medicine of Alexandria University. The risks and advantages of the procedure were explained to all patients and an informed consent was obtained from each of them.
Preoperatively, all patients had their detailed ocular and medical history taken. Full ophthalmic examination was performed including measurements of manifest refraction (MR), uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA) (using Snellen Decimal notation), and intraocular pressure as well as biomicroscopic fundus examination. Keratometric data, corneal topography, thickness data, and height maps of the anterior and posterior corneal surfaces were obtained from the Allegro Oculyzer
Following SMILE, postoperative follow-up visits were scheduled at 1 day, 1 week, and 1, 3, and 6 months postoperatively. During these visits, biomicroscopic examination of the anterior and posterior segments and UDVA, MR, and CDVA testing and recording were performed. Efficacy was expressed in terms of the cumulative UDVA at 6 months postoperatively as well as the efficacy index calculated as the ratio of the postoperative UDVA to the preoperative CDVA. Safety was judged by the change in the corrected distance visual acuity at 6 months postoperatively and also by the safety index calculated as the ratio of the postoperative CDVA to the preoperative CDVA.
Cases were operated by two surgeons (Osama Ibrahim and Tamer H. Massoud) in Roayah Vision Correction Centre, Alexandria, Egypt. Preoperatively, refractive data was fed and revised on the computer system linked to the VisuMax femtosecond laser system (Carl Zeiss Meditec AG, Germany) with a 500 kHz repetition rate. Data entered included the MR to be corrected (measured at 12 mm vertex distance), the mean corneal radius (mm), or mean
For all cases, small suction cups were chosen as the patient-laser interface. The cap and the lenticule diameters were calculated to be smaller than those of the graft (8.0 mm) so that they are centered within its margins. Corneal caps were planned to be 100
The carved lenticules had optical zones ranging from 5.5 to 6.0 mm based on the residual stromal depth which was always kept above 300
During surgery, the laser suction cup was centered relative to the pupil and the graft. The patient was asked to keep looking at the flickering green fixation light during laser application to the cornea. After creation of the cuts, the lenticule and the cap were manipulated using the usual techniques described for SMILE [
After extraction, the lenticule was spread on the corneal surface and stained with prednisolone acetate 1% eye drops to ensure its intactness as a complete disc and to detect the presence of any residual tissue remnants within the intrastromal pocket that can result in irregular astigmatism [
The same postoperative treatment regimen consisting of topical prednisolone acetate 1%, gatifloxacin 0.3%, and nonpreserved artificial tears was followed for all patients.
Data analysis was performed using the software SPSS for Windows version 20.0 (SPSS Inc., Chicago, USA) and Microsoft Excel 2010 (Microsoft corp., Redmond, WA, USA). Vector analysis for astigmatism results was done using Dr. Peyman’s astigmatic vector analyzer software (
Demographic and pre- and postoperative clinical (refractive and visual) data of the ten patients are shown in Table
Demographic, refractive, and visual data of postkeratoplasty smile patients.
Patient | Gender | Age | Manifest refraction | Visual acuity | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Preoperative | 6 months | Preoperative | 6 months | |||||||||
Sphere | Cylinder | Axis | Sphere | Cylinder | Axis | UDVA | CDVA | UDVA | CDVA | |||
1 | M | 28 | −4 | −3.5 | 110 | −0.5 | −0.75 | 115 | 0.2 | 0.8 | 0.7 | 0.9 |
2 | M | 31 | −5 | −1.5 | 25 | −1.75 | −1.5 | 40 | 0.1 | 0.8 | 0.7 | 0.8 |
3 | F | 27 | −7 | −4 | 60 | −2 | −1 | 25 | 0.05 | 0.6 | 0.6 | 0.7 |
4 | F | 25 | −6 | −3.25 | 10 | −1.75 | −1.75 | 30 | 0.1 | 0.5 | 0.4 | 0.6 |
5 | M | 33 | −7 | −2.5 | 35 | 0 | −2.5 | 70 | 0.1 | 0.7 | 0.7 | 0.9 |
6 | F | 35 | −5.5 | −2 | 70 | −0.5 | −0.5 | 90 | 0.1 | 0.8 | 0.8 | 0.9 |
7 | F | 25 | −4 | −2.5 | 65 | −1 | −0.5 | 60 | 0.3 | 0.9 | 0.8 | 0.9 |
8 | F | 28 | −4 | −2 | 15 | −1.5 | 0 | 0 | 0.2 | 0.8 | 0.8 | 0.8 |
9 | M | 27 | −5 | −6 | 105 | 1 | −2 | 115 | 0.05 | 0.5 | 0.8 | 0.8 |
10 | F | 31 | −5.5 | −3.5 | 165 | 0 | −1.25 | 180 | 0.1 | 0.9 | 0.5 | 0.9 |
M: male, F: female, UDVA: uncorrected distance visual acuity, and CDVA: corrected distance visual acuity.
In the majority of cases, the cap and the lenticule were centered in relation to the pupil within the corneal graft except in two cases (
Intraoperatively, some resistance was met during dissection of both the cap and the lenticule at the sites of the suture related fibrous tracks. However, this has not led to any complication and cases were completed as planned.
None of the cases enrolled in the study suffered from any of the reported post-SMILE complications.
The preoperative mean spherical error was −
The means of achieved correction values (calculated by subtracting the 6-month postoperative refraction from the preoperative target refraction) for sphere and MRSE were −
Vector analysis of the results of astigmatism correction revealed a mean target induced astigmatism (TIA) magnitude of
The mean correction ratio (induced/intended correction) of sphere was
Assessment of the accuracy of the achieved correction values versus the attempted ones revealed a positive correlation for all three refraction elements with the sphere showing correlation values of
Attempted versus achieved spherical correction 6 months postoperatively.
Target induced astigmatism versus surgically induced astigmatism 6 m postoperatively.
Attempted versus achieved spherical equivalent 6 m postoperatively.
As regards predictability, only one out of the ten eyes (10%) had a MRSE between zero and −0.5 D, 4 eyes (40%) were between zero and −1.0 D, and 7 eyes (70%) were between zero and −1.5 (Figure
Predictability of spherical equivalent correction 6 m postoperatively.
Figure
Stability of spherical equivalent refraction 6 m postoperatively.
On assessing the predictability of astigmatism correction, five out of the ten eyes (50%) had astigmatism above 3.0 D preoperatively, while six months after surgery 30% had astigmatism values within 0.5 D, 50% within 1.0 D, and 70% within 1.5 D (Figure
Predictability of refractive astigmatism correction 6 m postoperatively.
Preoperatively, the mean CDVA was
A more detailed evaluation of efficacy showed a cumulative Snellen preoperative CDVA of 0.9 or better in 20% of eyes, 0.8 or better in 60% of eyes, and 0.7 or better in 70% of eyes. On the other hand, the cumulative Snellen postoperative UDVA was 0.8 or better in 40% of eyes and 0.7 or better in 70% of eyes (Figure
Uncorrected distance visual acuity 6 m postoperatively.
Figure
Change in corrected distance visual acuity 6 m postoperatively (safety).
The quest for the best unaided visual performance following PKP has entailed the exploration of a variety of conventional and novel refractive correction procedures. Yet, none has proven enough refractive accuracy, predictability, efficacy, or safety to be adopted as the standard trustworthy technique. A wide variability as regards the obtained refractive and visual results in post-PKP eyes compared to results reported in healthy eyes with unoperated corneas has become a generally anticipated conclusion for all reports on such cases. In addition, the inherent complications of these refractive surgical correction techniques were found to have a higher incidence rate in grafted corneas adding an increased menace for the future viability of the graft and, rarely, the whole eye [
The introduction of the single step femtosecond laser small incision lenticule extraction (SMILE) for correction of myopia and astigmatism and its reported good results gave hope for a simple, fast, easily designed, readily centered, and theoretically safe refractive correction means that can be applied to corneal grafts while salvaging the circumferential graft-host interface scar as well as the endothelium from being violated [
This study was performed with the aim of evaluating visual and refractive outcomes after correcting postkeratoplasty myopia and myopic astigmatism using small incision lenticule extraction (SMILE).
To our knowledge, the only published data about SMILE after keratoplasty is a single case report in which the authors reported achieving the target refraction and an improved UDVA with a follow-up of 3 months. This study should, therefore, be one of the earliest clinical trials about the same subject [
The timing of intervention has been a matter of debate among researchers; however, it is generally agreed upon that the corneal graft-host junction heals completely about one year following transplantation and that further surgical interventions should not be done until three to four months has passed since all the sutures have been removed. Our cases had a minimum of 18 months before complete suture removal and refractive stabilization were pursued for three months afterwards [
As regards the refractive results, a statistically significant undercorrection was noted for sphere, cylinder, and MRSE. SMILE, in otherwise healthy nonoperated eyes, was reported to result in a slight undercorrection by about 0.25 D of MRSE as reported by Hjortdal et al. [
The statistically significant values of arithmetic and absolute angles of error of astigmatism correction denote the possibility cyclotorsion occurrence which can, also, aid the explanation of the relative imprecision of astigmatism correction at the intended angle as well as the induction of postoperative different axes cylindrical errors. Means for prevention of, compensating for, or correction of intraoperative cyclotorsion should be adopted.
Compared to other techniques, undercorrection and, rarely, overcorrection have also been reported for almost all corneal refractive surgical methods of correcting post-PKP myopia and astigmatism including incisional surgeries [
Despite this undercorrection reported for SMILE, stability of the achieved refraction was statistically proven on comparing the achieved MRSEs at 1 week, 1 month, 3 months, and six months. The same was reported by other studies which investigated SMILE for myopia and myopic astigmatism [
Our results also showed that SMILE for correcting post-PKP myopia and astigmatism is of high efficacy and safety. The value of the efficacy index (
The feasibility of centering the whole treatment within the graft had the advantage of avoiding the violation of the graft-host interface, thus, preserving the structural integrity of this potentially weak spot. On the contrary, in other nonfemtosecond laser dependent techniques, flap creation can easily breach (and consequently weakens) the circumferential scar [
The drawbacks of this study, however, include the absence of controls, the lack of randomization, the few number of the enrolled eyes, and multiple surgeons. Since higher order aberrations evaluation and visual quality assessment were beyond the scope of this study, we strongly advocate them to be done in future similar studies to ascertain the nature and the amount of induced higher order aberrations as well as the quality of vision provided following SMILE in grafted eyes. The need for more prolonged follow-up, evaluation of induced graft biomechanical changes, and comparison to other techniques used for the same indication cannot be overlooked.
To sum up, SMILE for correction of postkeratoplasty myopia and astigmatism can be considered a valuable addition to the armamentarium of procedures utilized to correct post-PKP myopia and astigmatism. It is effective, safe, and stable with moderate accuracy and predictability. The whole treatment can be centered within the graft preserving the viability of the healed graft-host interface. However, management of cyclotorsion as well as centration of both the graft initially and the lenticule afterwards is crucial for achieving the best refractive results.
The research protocol was revised and approved by the medical research ethical committee of Alexandria Faculty of Medicine.
An informed consent was obtained from the subjects after the nature of procedure had been explained.
No assistance with study design, data collection, data analysis, or manuscript preparation was obtained from any source. The protocol of the study was registered in Alexandria Faculty of Medicine research unit under the reference number 47/31 on 20/11/2014.
Tamer H. Massoud, Kitty Shehata, and Moones F. Abdalla declared no commercial or proprietary interest in the product or company and no financial interest as a consultant, reviewer, or evaluator and Osama Ibrahim is VisuMax research consultant for Carl Zeiss Meditec.