Collagen cross-linking (CXL) is a proven modality to stabilize kerectasia by improving stiffness of the cornea and halting progression of keratoconus.1,2 However, visual function does not improve much with this treatment alone, necessitating the need for additional refractive procedures.2 Thus, surgical techniques in combination with CXL have come into practice, leading to better visual outcomes, improving the quality of life, and possibly obviating the need for future corneal transplantation.
Topoguided photorefractive keratectomy (PRK) with either sequential or simultaneous CXL has been suggested to be safe and effective in long-term follow-up by regularizing corneal irregular astigmatism, decreasing aberrations and improving contact lens fit.3–7 Intracorneal ring segments (ICRs) in combination with CXL has also been found to improve refractive outcomes and stabilize early to moderate keratoconus.8–11 Toric (phakic) intraocular lenses after cross-linking have also been used with good success to correct residual refractive errors after CXL in patients not suitable for corneal correction.12
Laser in situ keratomileusis (LASIK) has conventionally been contraindicated in patients with keratoconus because of fear of ectasia. However, the latest flapless all-femtosecond procedure, refractive lenticule extraction (ReLEx), and its variant, small incision lenticule extraction (SMILE), when combined with accelerated CXL have provided a new ray of hope for individuals with early or forme fruste keratoconus; however, sufficient and long-term data on this new technique are lacking.13,14
Each available modality in the treatment of mild to moderate keratoconus combined with CXL seems to be feasible and working. Nevertheless, they have some limitations. PRK with CXL is a tissue subtractive technique, which although it improves refractive outcomes, further thins the keratoconic cornea, potentially increasing the risk of corneal destabilization. The cross-linking procedure when performed through the deepithelized surface increases the chances of haze, infection, and delayed recovery.15 ICRs, however, may be associated with potential risks of extrusion, misalignment, infection, perforation, and tissue reactions.16 Although tissue-saving techniques such conductive keratoplasty combined with CXL are described, they have not gained popularity because of the high rate of regression.17
In this era of femtosecond lasers and refractive lenticule technology, it may be possible to use cryopreserved extracted lenticules after ReLEx SMILE to serve as a tissue addition donor material. It was shown that cryopreservation does not cause deleterious effects on the collagen ultrastructure and hence may be potentially used to treat various conditions.18 The use of such cryopreserved lenticules to treat hyperopia with a tissue addition procedure has previously been described.19 We report another application of tissue addition for potential management of keratoconus in combination with accelerated CXL. A similar tissue addition technique using a stromal lenticule from a deceased donor followed by implantation into a femtolaser-enabled stromal pocket without simultaneous cross-linking for keratoconus has been reported previously, but long-term data on its safety and efficacy are lacking.20 In this prospective case series, we present our initial experience and 6 months of clinical outcomes with a new tissue addition technique that we call “femtosecond intrastromal lenticular implantation” (FILI).
The study was approved by the institutional ethics committee and adhered to the tenets of the Declaration of Helsinki. The trial was registered with Clinical Trial Registry of India (CTRI/2014/09/005012). Formal informed consent was obtained from the participants after they were explained the nature and possible benefits and risks of the procedure.
Inclusion criteria were grade 1 to 3 keratoconus21 with or without documented evidence of progression of keratoconus22; best spectacle-corrected visual acuity >20/200; age 21 to 40 years; intolerance to rigid gas-permeable, Rose K, or scleral lenses; central cone on topography (apex of the cone within a 3-mm zone); steep keratometry <65 D; minimum thinnest pachymetry of 400 μm; endothelial cell density (ECD) >1500 cells per square millimeter; willingness to sign informed consent; and participation in follow-up visits.
Eyes with advanced keratoconus with apical scarring, eccentric cone distribution on topography, axial myopia (axial length >24 mm), active atopic keratoconjunctivitis, corneal ectasia due to previous laser ablative surgery (LASIK/PRK), previously failed CXL, history of riboflavin allergy, moderate to severe dry eye, concurrent use of corticosteroids or antimetabolites, pregnant or breastfeeding mothers, and patients with unrealistic expectations were excluded from the study.
Baseline clinical examination included evaluation of uncorrected visual acuity (UCVA), best spectacle- or contact lens-corrected distance visual acuity (CDVA), slit-lamp biomicroscopy, dilated fundus examination, corneal topography with the Orbscan IIz (Bausch & Lomb, NY) and Sirius (SCHWIND eye-tech-solutions, Kleinostheim, Germany), Anterior Segment Optical Coherence Tomography (AS-OCT) (Optovue, CA) for the epithelial thickness profile, Corvis ST (Oculus Optikgeräte GmbH, Wetzlar, Germany) for corneal biomechanics, specular microscopy (Tomey, Japan) for ECD, and aberrometry (iTrace; Tracey Technologies, TX) for higher-order aberrations and coma aberrations.
For the treatment of keratoconus, cryopreserved lenticules of known diameter and thickness obtained after ReLEx SMILE for myopia correction were used. ReLEx SMILE was first performed on eligible patients with myopia using the Visumax femtosecond laser (Carl Zeiss Meditec AG, Jena, Germany) with a pulse repetition rate of 500 kHz, energy cut index of 33 nJ, and a spot distance of 4.5 μm. The lenticule diameter (optical zone) was 6.0 to 7.0 mm, side-cut angle 90 degrees, cap diameter 7.5 mm, cap thickness of 120 μm, and a superior 2-mm incision.
Lenticules extracted after the procedure were cryopreserved after obtaining informed consent from donors and screening the donors for the presence of human immunodeficiency virus, hepatitis B, and hepatitis C. Only lenticules with spherical myopic refractive errors were chosen for cryopreservation. These lenticules were subjected to a modified tissue-processing technique and cryopreserved. The details of the methods have been described earlier.19
We attempted to use elevation topography as a guide to calculate the required amount of tissue to be added for the desired effect, but it did not seem to be feasible because of irregular topography and variability of the best-fit sphere in the ectatic area, which made the calculations too complex. Hence, for simplicity, we selected lenticules on the basis of spherical equivalent refraction of the recipient eye after ruling out axial myopia, because our aim was mainly to study the flattening and refractive effects of tissue addition and its safety when combined with accelerated CXL in keratoconus.
After thawing and washing the cryopreserved lenticule with a previously described technique,19 the tissue was placed on a Teflon block with the anterior aspect of the lenticule facing up. The side of the lenticule was identified by a 10-0 nylon suture that was tied on the superior side of the lenticule at the time of lenticule collection. This suture was cut; the surface of the lenticule was cleaned using 2 Weck-Cel sponges (Beaver Visitec, MA) and dried while centering it on the central black mark of the Teflon block. Once the tissue was well centered, 2 to 3 drops of 0.25% riboflavin dye (VibexXtra; Avedro, MA) were applied to the surface of the tissue to enhance the visibility of tissue against the white background of the Teflon block. Excess dye was removed with a Weck-Cel sponge (Beaver Visitec), and the surface dried. A 3-mm corneal trephine was then used to punch the center of the lenticule under an operating microscope while ensuring a uniform width of frill of the tissue around the trephine, resulting in a donut-shaped tissue. This tissue was soaked in riboflavin dye and kept aside (see Video, Supplement Digital Content 1, https://links.lww.com/ICO/A291).
Under topical anesthesia, the Visumax femtosecond laser (Carl Zeiss Meditec AG) with an energy cut index of 35 nJ and a spot distance of 5 μm was used to create a pocket into the patient's cornea at a 100-μm depth with 7.0- to 8.0-mm diameter (1 mm larger than the optical zone of the donor lenticule) and a 4-mm superior incision. The pocket was dissected using a blunt spatula followed by injection of 0.25% VibexXtra (Avedro) dye into the interface for 60 seconds after which the interface was washed with normal saline. The corneal vertex was marked with gentian violet using the first Purkinje image as a reference while asking the patient to fixate on the microscope light. The donut-shaped lenticule with the anterior aspect facing upward was held with lenticule forceps and gently inserted into the pocket through the 4-mm superior incision. The lenticule was positioned around the marked center of the cornea and ironed out from the surface using a blunt spatula. Once centration was ensured, the wound was mopped with a Weck-Cel sponge (Beaver Visitec) to soak excess fluid from the interface. The eye was then exposed to UV-A radiation using the Avedro CXL system (Avedro), at 30 mW/cm2 for 3.3 minutes, delivering a total dose of 6.3 J/cm2 to the corneal surface. Postoperatively, antibiotic eye drops (Exocin; Allergan, CA) were prescribed 4 times for 1 week, and steroid eye drops (Pred Forte; Allergan) were prescribed with a tapering dose for 3 months, along with lubricating drops. Follow-ups of patients were conducted for a mean period of 190 ± 13 days (range, 117–193 days).
Postoperative examinations were performed on day 1, day 15, 3 and 6 months, which included assessment of UCVA, CDVA, retinoscopy, slit-lamp examination, topography with the Orbscan (Bausch & Lomb) and Sirius (SCHWIND eye-tech-solutions), AS-OCT (Optovue), Corvis ST (Oculus Optikgeräte GmbH), specular microscopy, aberrometry, and clinical photography after dilation for centration and stability. Postoperative corneal haze and folds were graded at every follow-up time point using the scale described by Nakamura et al23 and the Corneal Folds Grading Atlas.24
There were no intraoperative or postoperative complications at the end of each examination. All patients tolerated the procedure well. Slit-lamp examination on day 1 showed mild interface haze and few lenticular tissue folds, which had subsequently cleared by day 15. No reaction, infection, epithelial defects, punctate keratitis, deep lamellar keratitis, or signs of allogeneic rejection were observed in any of the treated eyes by the end of the follow-up period. Shift was documented using serial digital photography on all postoperative visits. A line was drawn from the edge of the lenticule to the limbus at 3 reference points. On each postoperative visit, this distance was measured and compared with previous photographs. This distance was found to be constant, which indicated a stable position of the lenticule over time. Figures 1A, B show clinical images at 1 month and 6 months of an eye treated with FILI + CXL exhibiting good lenticule stability. Figure 1C shows AS-OCT of the same eye at 6 months postoperatively.
FIGURE 1:A and B, Clinical images of the right eye of a patient treated with FILI + CXL at day 15 and 6 months postoperatively, showing a clear and stable donut-shaped lenticule over time. C, AS-OCT image of the same eye at 6 months showing a clear and well-centered lenticule and a stromal demarcation line at 211 μm.
Six eyes of 6 patients (5 men and 1 woman) with a mean age of 19.5 ± 3.1 were treated with this technique. The preoperative data of recipient patients and details of lenticules implanted are shown in Table 1. The mean length of cryopreservation of lenticules used in the study was 84.5 ± 70.2 days.
TABLE 1:Preoperative Data of Recipients and Details of the Lenticules for the Eyes Treated With FILI + CXL (n = 6)
All 6 eyes showed improvement in uncorrected visual acuity, whereas 4 eyes showed improvement in corrected visual acuity. Two eyes gained 4 lines, 1 eye gained 3 lines, 1 eye gained 2 lines, whereas 2 eyes had no change in lines of CDVA (Table 2). This was associated with reduction in spherical equivalent and astigmatism postoperatively in all eyes. No eye lost lines of CDVA.
TABLE 2:Visual and Refractive Results of (n = 6) Eyes at 6 Months Follow-up
All eyes revealed generalized flattening of mean keratometry on Sirius topography. The flattening was clinically more apparent in 3-mm and 5-mm zones than in the 7-mm zone (Table 3). Figure 2 shows a difference map of anterior tangential keratometry and anterior elevation at 6 months postoperatively of an eye treated with FILI + CXL on Sirius (Schwind eye-tech-solutions) topography. All eyes revealed reduction in the anterior Q value at 6 months, suggesting a more regular and prolate corneal surface (Table 3).
TABLE 3:Changes in Keratometry and Asphericity (Q Value) at 6 Months Postoperatively for (n = 6) Eyes Treated With FILI + CXL on the Sirius
FIGURE 2:Difference map of anterior tangential keratometry and anterior elevation with the Sirius for an eye preoperatively and at 6 months postoperatively.
In all 6 eyes, both central corneal thickness and midperipheral thickness (corresponding to a 5-mm zone) were seen to be increased postoperatively. The change was more evident in the midperipheral zone than in the central zone, as expected (Table 4).
TABLE 4:AS-OCT Changes in Pachymetry and the Epithelial Thickness Profile in Central and Midperipheral (5 mm) Zones at 6 Month postoperatively for (n = 6) eyes
A corresponding increase in epithelial thickness was also seen, although the change was more apparent in the central zone than in the midperiphery (Table 4). The epithelial thickness measurements obtained were within the limits of measurement errors.
Figure 3 shows the comparison of epithelial thickness maps before and 6 months after the procedure. The stromal demarcation line25 on AS-OCT was seen in 5 eyes as early as 15 days and up to 6 months in these eyes, whereas in 1 eye, the line was not clearly appreciated at any time point postoperatively. The average depth at 3 months was at 210 ± 11 μm.
FIGURE 3:Comparison of the epithelial thickness profile for an eye treated with FILI + CXL preoperatively and at 6 months postoperatively.
There was a decrease in mean root mean square values of higher-order aberrations and coma aberrations postoperatively in all eyes. Mean endothelial cell counts remained stable at 6 months postoperatively (see Table 1, Supplemental Digital Content 2, https://links.lww.com/ICO/A292).
Using the Corvis ST Scheimpflug noncontact tonometer, intraocular pressure (IOP), A1 (length of first applanation), A2 (length of second applanation), deformation amplitude, peak distance, and R (radius) at the highest concavity were studied and compared with preoperative values. Clinically, the mean values for IOP, A1, A2, and R were seen to be increased, whereas the mean deformation amplitude and peak distance decreased at 6 months postoperatively (see Table 2, Supplemental Digital Content 3, https://links.lww.com/ICO/A293).
Tissue addition techniques to change the refractive error of the eye have been previously described. Keratophakia using either cryolathed homograft discs or factory-lathed hydrogel lenticules, which are inserted between corneal lamellae separated by a microkeratome has been reported as a feasible method to treat myopia, hyperopia, and aphakia by changing the anterior curvature of the cornea.26,27 However, it did not become popular because of the lack of refractive predictability and technical difficulties. Complications such as extrusion, erosion through the epithelium, and encapsulation of the lenticule by activated keratocytes were also reported with hydrogel implants.28 Epikeratophakia overcame many of the objections to earlier lamellar refractive procedures, such as the lamellar dissection of the recipient's healthy cornea over the visual axis and the need for expensive surgical equipment. Apart from other refractive errors, it could also be used to treat keratoconus. It was reversible and repeatable, and serious and visually threatening complications were less. However, the lack of precision during manufacturing of the lenticules leading to loss of the refractive effect and suture-related complications were the limitations because of which this technique went into disrepute.29 Recently, use of the MyoRing (Dioptex, Linz, Austria) flexible full-ring implant, which is inserted into a corneal pocket with a high-precision microkeratome (PocketMaker; Dioptex), has also been suggested to allow safe, easy, and effective treatment of myopia, keratoconus, and post-LASIK keratectasia.30
The use of cryopreserved refractive lenticules extracted after ReLEx SMILE in subjects with FILI (femtosecond intrastromal lenticule implantation) for correction of hyperopia has been recently reported.19 It was proposed that this technique, when combined with accelerated CXL, may be used for potential treatment of mild to moderate keratoconus in this exploratory study.
The aims of keratoconus treatment are to stabilize the cornea and modify the corneal shape to improve its optics and quality of vision by reducing aberrations. Also, reduction in associated myopic astigmatism is desirable because most of these patients are diagnosed in refractive clinics.
The rationale of using this technique for treating keratoconus is addition of corneal tissue in the midperiphery and around the cone to cause relative flattening in the center, along with a modification in the corneal shape to a more natural prolate one. Accelerated CXL is performed to aid in stabilizing the final corneal shape to prevent further progression of keratoconus. Previous studies have shown comparable clinical outcomes and a reduced treatment time with accelerated CXL when compared with conventional cross-linking.31 Hence, we preferred accelerated CXL in this procedure, which was facilitated by injection of the dye into a corneal pocket in contrast to conventional cross-linking, which involves removal of the epithelium. The safety and efficacy of CXL in the femtolaser-enabled pocket has been suggested by Kanellopoulos.32 This ensures no or less haze, faster healing, better diffusion of the dye, no risk of infections, and more patient comfort.
The stromal demarcation line after accelerated CXL in our series was seen approximately at 211 μm, which is deeper compared with the level of pocket creation (100 μm) and riboflavin injection. However, our observations are consistent with other studies on accelerated CXL that have also shown the demarcation line to be more superficial but equal in efficacy compared to conventional CXL.31
It is debatable whether this technique is analogous to the ICRs in terms of the mechanism of action because both techniques involve addition of the material in the midperiphery. Although ICRs are polymethylmethacrylate rings that act as a spacer between the corneal lamellae causing shortening of the central arc proportional to the ring thickness,33 we theorize that FILI involves addition of natural corneal tissue, which acts more like a “filler” and seems to cause local elevation in the midperiphery and relative flattening in the center, without actually causing much tension or pull on the corneal lamellae as the ICRs do.
In addition, the depth at which both are implanted differs. ICRs are inserted at 80% of the corneal depth, whereas in our technique, we inserted the lenticule more anteriorly at a fixed depth of 100 μm from the epithelial surface. We chose a depth of 100 μm to obtain a maximum anterior effect of tissue addition. Our experience with tissue addition for hyperopia suggests that implanting tissue at the depth of 160 μm causes the posterior cornea to push posteriorly, creating changes in Descemet membrane, especially with thicker lenticules. Second, 100 μm is the ideal depth for CXL for effective cross-linking when performed through a corneal pocket.32
Hence, although both techniques involve addition of the material in midperiphery, the outcomes may differ. Therefore, nomograms used for ICRs based on the arc length, diameter, and thickness of the ring may not apply for FILI, although the factors considered for surgical planning such as the grade of keratoconus, type of the conus, and degree of myopia and astigmatism (spherical equivalent) are essentially the same.
In this series, 5 eyes with low to moderate progressive keratoconus (mean keratometry <58 D) and 1 eye with advanced progressive keratoconus (eye 4, mean keratometry 64 D, devoid of apical scarring) with central/symmetrical cones were treated. The eye with advanced keratoconus did not show reduction in mean keratometry after operation unlike other eyes, which suggests that this technique may not be favorable in advanced stages of keratoconus with very steep corneas (K > 58 D). In such advanced cases and ones with central scarring, lamellar or full-thickness transplant may be the only valid option.
The major benefits of this technique lie in the change in the shape of a keratoconic cornea, which is invariably hyperprolate to a more favorable prolate/less hyperprolate shape. In all the treated eyes, a significant reduction in the Q value was noticed after tissue addition, irrespective of the preoperative keratometry, pachymetry, and refractive error. A corresponding decrease in both higher-order and coma aberrations was seen in all eyes. Both uncorrected and best-corrected visual acuities improved and were maintained until the last follow-up.
Of all the noncontact methods available, both AS-OCT and Scheimpflug imaging provide equally reproducible measurements for corneal thickness assessment after laser refractive surgery, although the accuracy has been observed to be better with AS-OCT.34 Hence, we preferred AS-OCT over the Sirius to analyze corneal thickness changes. Our observations revealed an increase in both central and midperipheral pachymetry, which remained stable over the follow-up period. A possible cause of the increase in central pachymetry may be mild lifting of the most anterior corneal layers and creation of a potential space in the center after tissue addition in the midperiphery. Interestingly, we observed that central pachymetry showed a trend toward a slight decrease with time (immediate postoperative through 6 months), possibly suggesting that the “potential space” created in the center after tissue addition is collapsing or is being filled with an extracellular material. Long-term follow-up is required to establish this theory. Also, compensatory epithelial hyperplasia due to flattening of the central curvature also seems to have contributed to the increase in central corneal thickness postoperatively.
The epithelial thickness profile in keratoconus shows a typical donut pattern characterized by a localized central zone of thinning overlying the cone, surrounded by an annulus of thick epithelium.35 In our series, epithelial thickness was seen to be increased in the central zone (corresponding to the apex of the cone), whereas it was not so in the midperipheral zone of lenticule implantation suggesting a more regular pattern of epithelial thickness after tissue addition.36
Corneal biomechanics after CXL is a controversial topic, and it has been suggested that instruments based on fast deformation of the cornea, such as the Corvis ST and Ocular Response Analyzer, may be too insensitive to the changes in biomechanics induced by corneal cross-linking.37 We noticed improvement in some parameters such as the deflection amplitude after the procedure, earlier stated to be linked to corneal biomechanics,38 but based on the small number of eyes, it is difficult to draw any conclusion regarding effects of tissue addition and accelerated CXL; this area needs further research.
Initial experience with the small number of eyes suggests that FILI with CXL may be an alternative option to treat mild to moderate grades of keratoconus. It is too early to comment whether this can be a substitute for ICRs. Nevertheless, it seems to be a feasible and safer approach because complications such as extrusion, migration, perforation, asymmetrical placement, chronic pain, explantation, keratitis, and corneal melt reported with ICRs are less expected. Also, there is no further thinning of the cornea as with topoguided PRK.
The technique is still in its infancy and needs a longer follow-up period to answer the concerns about regression due to tissue modulation, maintenance of stabilizing effects, and allogeneic graft rejection. Also, its efficacy in halting the progression of keratoconus when used in combination with accelerated CXL needs to have a longer follow-up period. Nomograms need to be refined for better predictability of refractive results and wider application. Nevertheless, to our knowledge, this is the first report on the use of cryopreserved corneal lenticules for the treatment of keratoconus, which indicates encouraging results that could possibly change current practice in the management of keratoconus. However, as this article mainly aims at reporting preliminary outcomes with this new technique, longer follow-up is required to establish our results in terms of stability of the cornea and refractive outcomes.
With high quality products and considerate service, we will work together with you to enhance your business and improve the efficiency. Please don't hesitate to contact us to get more details of custom lenticular patch.