BASIC INVESTIGATION
Riboflavin–UV-A Crosslinking for Fixation of Biosynthetic Corneal Collagen Implants Kerstin Wand, MD,* Raphael Neuhann,* Andrea Ullmann,* Katharina Plank,* Michael Baumann,† Roland Ritter,‡ May Griffith, Prof,§ Chris P. Lohmann, MD,* and Karin Kobuch, MD*
Purpose: To evaluate riboflavin–UV-A crosslinking as an alterna-
be further evaluated. Biostability, integration, and long-term outcome are further evaluated in in vivo animal experiments.
Methods: A range of cell-free corneal implants consisting of
Key Words: collagen implant, riboflavin–UV-A crosslinking, biosynthetic cornea, laser DALK, corneal transplantation, keratoprosthesis
tive suture-free fixation method for biosynthetic corneal collagen implants. recombinant human collagen type III were examined. In vitro, the implants were crosslinked with different riboflavin solutions and irradiations. Ex vivo, the biosynthetic corneal implants were placed on the anterior cornea of porcine and rabbit eyes after performing deep anterior lamellar keratoplasty with a trephine, femtosecond laser, or excimer laser. UV-A crosslinking was performed with isotonic or hypotonic riboflavin at either standard or rapid procedure. The corneas were excised, fixed in PFA 4%, and embedded in paraffin. Crosslinking effects on the implants and the adhesion between implant and corneal bed were evaluated by slit-lamp biomicroscopy, optical coherence tomography (OCT) images, and histologically.
Results: After the crosslinking procedure, the implants showed different degrees of thinning. The accuracy of cutting the corneal bed was highest with the excimer laser. Good adhesion of the implant in the corneal bed could be demonstrated in OCT images. This was more accurate in porcine eyes than in rabbit eyes. Histologically, crosslinks between implant and corneal stroma were demonstrated. There was no difference between standard and rapid crosslinking procedures.
Conclusions: Riboflavin–UV-A crosslinking as a fixation method for biosynthetic corneal collagen implants was demonstrated to be promising. It can reduce suture-related complications such as haze formation and surface irregularity. Stability of the implants, especially shrinkage after riboflavin–UV-A crosslinking, needs to Received for publication October 17, 2014; revision received January 4, 2015; accepted January 19, 2015. Published online ahead of print March 13, 2015. From the *Department of Ophthalmology, Klinikum rechts der Isar, Technische Universität München, Munich, Germany; †MLase AG, Germering München, Munich, Germany; ‡Technolas PV Bausch & Lomb, Munich, Germany; and §Department of Clinical and Experimental Medicine and Integrative Regenerative Medicine Centre, Linköping University, Linköping, Sweden. Supported by EU nanomedicine ERA-net project I-CARE, with funding from VDI (Verein Deutscher Ingenieure) and VDO to MLase AG, Swedish Research Council to MG. M. Baumann works for MLase AG; R. Ritter works for Technolas PV Bausch & Lomb. The remaining authors have no conflicts of interest to disclose. Reprints: Kerstin Wand, MD, Department of Ophthalmology, Klinikum rechts der Isar, Technische Universität München, Ismaningerstr. 22, D-81675 München, Germany (e-mail: [email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
544
| www.corneajrnl.com
(Cornea 2015;34:544–549)
C
orneal disease or damage leading to loss of corneal clarity is the ninth largest cause of permanent loss of vision or blindness worldwide.1 Currently, the most successful treatment for blindness due to corneal opacification is the transplantation of human donor corneal tissue, either by full penetrating keratoplasty or partial lamellar keratoplasty. However, grafting with donor tissue depends on the available supply of donor corneas. Because of the current shortage of acceptable donor corneas, there are long waiting lists of patients to receive a transplant. In the future, demographic age changes and an increasing number of refractive corneal surgeries will further reduce the amount of possible donor corneas. The current alternative to human donor tissues is artificial corneas or keratoprostheses, such as the Boston keratoprosthesis, the most commonly implanted artificial cornea.2 However, because of the possible severe complications and lack of host–graft integration, keratoprostheses are currently restricted to patients not suitable for donor tissue grafting, for example, in case of repeated allogeneic tissue failure or severe ocular surface diseases.3 Recently developed biosynthetic corneal implants could become a primary alternative to allogeneic transplants.4 The cornea is transparent, avascular, and immunologically privileged, which makes it an excellent candidate for tissue engineering for transplantation.5 Cell-free corneal-shaped matrices, based on crosslinked collagen, were introduced for corneal transplantation.5–8 Biosynthetic corneal implants have been applied in multiple in vivo animal experiments and showed an overall good integration in donor tissue, maintenance of optical clarity, epithelial coverage of the implants, cell and nerve ingrowth from donor tissue, and restored touch sensitivity.5–9 A phase 1 clinical study of implanted corneal substitutes (composed of EDC crosslinked recombinant human collagen type III) was conducted on 10 patients to replace the pathological anterior cornea.10 After anterior lamellar keratoplasty (ALK), the biosynthetic corneal substitutes were implanted with overlying mattress sutures that were removed after 6.5 weeks. Overall, the implants showed Cornea Volume 34, Number 5, May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
good integration in donor tissue without rejection, vascularization, infection, or the need for long-term steroid immunosuppression for more than 4 years.11 The implants were covered by stable morphologically normal epithelium, tear production and osmolarity were normal, stromal cell migration from the recipient into the implant occurred, regenerating nerves were noted after 3 months of operation, and corneal touch sensitivity returned within the first 12 months after the operation. However, suture-associated complications were noted. Areas of focal haze with thinning of the implant were found in the central cornea that corresponded to the intersection point of the overlying mattress sutures.10 The presence of tight overlying sutures had delayed epithelialization in these areas, creating an epithelial defect and subsequently an inflammatory response leading to implant thinning and invasion of repair fibroblasts. The localized thinning induced surface irregularity that was noticed postoperatively and required compensation by placement of rigid contact lenses for best-corrected visual acuity. A suture-free retention method such as riboflavin–UV-A crosslinking would therefore be expected to improve the visual outcome. Corneal crosslinking induced by riboflavin–UV-A is currently accepted for treatment for keratoconus. In numerous studies, it was shown that riboflavin–UV-A crosslinking increases long-term corneal biomechanical stability and its resistance toward collagenases.12–15 An increase in collagen fiber diameter in the anterior stroma of the cornea is assumed to be the morphological correlate to the increase of biomechanical stability.16 The biochemical background is explained by the production of oxygen radicals induced by the photosensitizer riboflavin in combination with UV-A. Oxygen radicals interfere with the amino acids of collagen molecules and this photochemical process leads to intrafibrillar and interfibrillar covalent bonds in collagen fibrils.17 In patients with keratoconus, riboflavin–UV-A corneal crosslinking has been proven to prevent the disease from progression by stabilizing corneal keratometry.18–23 It is a simple and safe procedure as long as the cytotoxic threshold of irradiance is not exceeded and the endothelium is not damaged.24–26 In this study, our purpose was to establish a new suturefree retention method for biosynthetic corneal collagen implants to resolve suture-related complications such as thinning of the implant and delay in epithelialization leading
Riboflavin–UV-A Crosslinking
to haze and surface irregularity. Using its photochemical effect of forming covalent bonds in collagen fibrils, riboflavin–UV-A corneal crosslinking could be used as an excellent tool for fixation of biosynthetic corneal implants in the recipient corneal stromal bed.
METHODS Different cell-free collagen hydrogels were used as biosynthetic corneal implants. The hydrogels were fabricated from recombinant human collagen type III (RCH III; Fibrogen Inc, San Francisco, CA) crosslinked with 1-ethyl-3-(3dimethylaminopropyl) carbodiimide and N-hydroxysuccinimide (EDC/NHS; Sigma-Aldrich, St Louis, MO). Several of these RHCIII implants were reinforced with a second network comprising 2-methacryloyloxyethyl phosphorylcholine (MPC; Biocompatibles, UK and Paramount Fine Chemicals Co Ltd, Dalian, China). Fabrication of these interpenetrating networks have been described in detail in previous publications.8,27,28 The RHCIII substitutes consisted of different RHCIII concentrations, 10, 13.7, 15, and 18%. The RHCIII–MPC substitutes consisted of RHCIII 13.7% mixed with MPC. The resulting implants were corneal shaped, 350 mm in thickness, and 12 mm in diameter. For ex vivo implantation, the corneal substitutes were either trephined to obtain 6 mm diameter buttons (Fig. 1A), or cut into shapes that complemented that of the recipient corneal bed using either an excimer laser (MLase AG, Munich, Germany) or a femtosecond laser (Technolas PV, Munich, Germany). We first ascertained that collagen hydrogels could be crosslinked with riboflavin–UV-A. RHCIII hydrogels with starting collagen concentrations of 10, 13.7, 15, and 18% were divided into 4 subgroups of 3 implants. The first group was UV-A crosslinked with isotonic riboflavin 0.1% (10 mg riboflavin-5-phosphate in 10 mL dextran T500 20% solution), the second group with hypotonic riboflavin (0.1% riboflavin in hypoosmolar solution). Riboflavin solution was applied every minute for 5 minutes before UV-A irradiation was started using the rapid protocol (18 mW/cm2 for 5 minutes) (CCL 365; MLase AG). The third and fourth groups were controls comprising implants that were treated with either isotonic or hypotonic riboflavin solution for 5 minutes only without any subsequent UV-A irradiation. Implant thickness
FIGURE 1. A, Preparation of the implant by trephining out a button of the corneal substitute. B, Corneal bed in porcine eye after deep anterior lamellar keratoplasty with excimer laser. C, Riboflavin–UV-A crosslinking of implant embedded in the corneal bed. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.corneajrnl.com |
545
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
Wand et al
was assessed before and after treatment using OCT imaging (iVue 100; Optovue Europe GmbH). Biosynthetic corneal implants (20 RHCIII hydrogels of 13.7% and 15 RHCIII–MPC hydrogels) were “implanted” into 20 ex vivo porcine eyes and 15 ex vivo rabbit eyes. The isolated globes underwent deep anterior lamellar keratoplasty ablation. Wound beds of 6 mm in diameter and 350 mm in thickness were cut into 6 porcine eyes and 250 mm in thickness into 5 rabbit eyes by handheld corneal trephine. An excimer laser (MLase AG) was used to make analog cuts in an additional 7 porcine and 5 rabbit eyes. A femtosecond laser was used on a further 7 porcine and 5 rabbit eyes (Technolas PV Bausch & Lomb) (Fig. 1B). For crosslinking, each corneal implant was positioned within its target corneal bed and riboflavin–UV-A corneal crosslinking was performed (Fig. 1C). Isotonic riboflavin 0.1% solution was applied on the contact side of the implant and stromal corneal bed for 5 minutes, allowing the riboflavin to permeate through the implant and cornea. Then, UV-A irradiation was started at either standard or rapid protocol. Directly after corneal crosslinking procedure, the implanted eyes were examined by slit-lamp biomicroscopy and OCT images (iVue 100; Optovue Europe GmbH). The corneas were then excised manually by circumferential scleral incision, fixed in paraformaldehyde 4%, and embedded in paraffin for subsequent histological examination after hematoxylin and eosin (H&E) and picrosirius red staining, as well as scanning electron microscopy (SEM). Specimens with implants treated with riboflavin but without UV-A crosslinking served as controls.
TABLE 1. Change in Thickness of RHCIII Implants After Treatment With Riboflavin Solution (Isotonic or Hypotonic), With or Without UV-A Irradiation (Rapid Protocol) Demonstrates Different Degree of Shrinkage of the Implants RHCIII 10%
13.7%
15%
18%
Average initial thickness of implant 316.41 371.41 350.75 330.90 before treatment, mm Change in thickness after treatment 253.07 236.10 228.52 237.00 with isotonic riboflavin solution, % Change in thickness after treatment 16.30 22.86 2.01 4.00 with hypotonic riboflavin solution, % Change in thickness after treatment 257.5 255.5 244.95 250.18 with isotonic riboflavin solution + UV-A irradiation, % Change in thickness after treatment +2.3 220.6 26.76 28.69 with hypotonic riboflavin solution + UV-A irradiation, %
The biosynthetic corneal implants consisting of EDC/ NHS crosslinked RHCIII of different concentrations showed different degrees of shrinkage after treatment with riboflavin solution (isotonic and hypotonic) and after riboflavin–UV-A crosslinking. Table 1 shows the average change in implant thickness (measured in micrometers in OCT images) in the 4 subgroups of each RHCIII concentration (10, 13.7, 15, and 18%) after treatment with isotonic or hypotonic riboflavin solution solely and after treatment with riboflavin solution (isotonic or hypotonic) followed by UV-A irradiation (rapid protocol, 18 mW/cm2 for 5 minutes). Table 1 demonstrates that treatment with isotonic riboflavin solution solely leads to a significant change in the RHCIII implant thickness on average by 39% compared with their initial value. Treatment with hypotonic riboflavin solution in contrast did not reduce implant thickness significantly. Treatment with isotonic riboflavin solution followed by UV-A irradiation (rapid protocol) showed an average reduction of implant thickness by 52% compared with its initial value. Treatment with hypotonic riboflavin solution followed by UV-A irradiation in contrast reduced the implant thickness on average by only 8%. Higher collagen content of RHCIII seemed to reduce the effect of riboflavin osmolarity but caused an additional shrinkage by UV-A irradiation and crosslinking (Table 1). RHCIII implants (13.7% collagen) and RHCIII–MPC implants, respectively, were implanted into ex vivo porcine and rabbit eyes. Corneal beds were prepared using 1 of the
following: a 6-mm handheld corneal trephine, an excimer laser, or a femtosecond laser. There was a noticeable difference in the surface quality of the corneal bed that was dependent on the device that was used for preparation. The excimer laser produced the smoothest surfaces with few to no irregularities. The femtosecond laser and handheld trephine produced a rather rough surface with little bumps and holes that were visible in OCT images, which led to greater difficulty in the adhesion of the implants. Riboflavin–UV-A crosslinking of the implants lying in the corneal bed resulted in permanent adhesion. The implants stayed adherent in the corneal bed for all further investigations. A scheme of the “ideal apposition” between implant and corneal bed for better understanding of the histological and OCT images is demonstrated in Figures 2A, B. Slit-lamp biomicroscopy showed tight adhesion between implant and corneal bed after the crosslinking procedure (Fig. 2C). OCT images confirmed the apposition between implant and corneal bed (Figs. 2D, E, arrowheads point at interface). After excision and fixation of the corneas, the connection between implant and corneal bed could also be demonstrated histologically (in H&E-stained and picrosirius-stained sections; Fig. 2F). In this histological section, due to implant thinning, the implant lies deep in the corneal bed and the surface of the implant and corneal recipient is not on a plane (Fig. 2F). The changes in implant thickness after riboflavin–UV-A crosslinking seen in our in vitro experiments were partially also observed in the ex vivo experiments in porcine and rabbit eyes. On higher magnification using SEM, direct fiber links between implant and corneal stroma could be found at their interface (arrow, Fig. 2G). The fitting of the implant was more accurate in porcine eyes than in rabbit eyes, and better integration could be demonstrated after deep anterior lamellar keratoplasty with excimer laser than with femtosecond laser or handheld trephine. In contrast, in the control group, in which the implants were appositioned within the surgical bed but without riboflavin–UV-A crosslinking, the implants detached completely. No further examination was possible.
546
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
RESULTS
| www.corneajrnl.com
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
Riboflavin–UV-A Crosslinking
FIGURE 2. A, Scheme of “ideal apposition” between implant and corneal bed as in OCT images. B, Scheme of “ideal adhesion” between implant and corneal stroma as in histological sections. C, Macroscopic image of RHCIII implant in a prepared corneal bed. D, Implant in porcine corneal bed after deep lamellar keratoplasty with excimer laser in OCT image (note that layered appearance of the RHCIII–MPC implant). The smooth graft–host interface is marked by arrowheads. E, Implant in rabbit corneal bed after deep lamellar keratoplasty with femtosecond laser in OCT image. The bumpy graft–host interface is marked by arrowheads. F, Good adhesion of implant to the host corneal stroma is seen in this H&E-stained section. G, Scanning electron microscopy showing physical structural connections (arrow) between implant (parallel orientation of collagen fibers) and corneal stroma.
Overall, there was no difference in the outcome between UV-A irradiation with standard (3 mW/cm2 for 30 minutes) and rapid protocols (18 mW/cm2 for 5 minutes).
DISCUSSION Biosynthetic corneal implants could become a primary alternative to transplantation of human corneal donor tissue in the treatment of corneal opacification, which would overcome the existing shortage of acceptable donor tissues.4 Multiple in vivo animal experiments and a phase 1 clinical study showed good results overall with regard to integration and optical clarity.5–11 However, there were suture-associated complications such as thinning of the implants and delayed epithelialization at earlier suture points leading to surface irregularity and haze formation.10 Therefore, a new suture-free Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
retention method would be expected to improve the visual outcome. The purpose of this study was to establish and assess an alternative suture-free fixation method using riboflavin–UVA crosslinking for incorporation of biosynthetic corneal implants in the recipient corneal stromal bed. In our in vitro experiments, the biosynthetic implants showed different degrees of shrinkage after treatment with riboflavin solution solely or riboflavin solution plus UV-A irradiation. Treatment with isotonic riboflavin solution led to significant thinning of the implants. However, treatment with hypotonic riboflavin solution did not alter the thickness, suggesting an osmotic reaction. It should be noted that the RHCIII implant comprises 85% to 90% water; so, the isotonic riboflavin solution with 20% dextran would be hyperosmolar compared with the implant, resulting in thinning. The hypotonic solution, however, is mainly water, just like the www.corneajrnl.com |
547
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Wand et al
Cornea Volume 34, Number 5, May 2015
implants and hence there was less shrinkage. Modulation of corneal thickness by various riboflavin eye drops compositions has been described previously and confirms our observation.27 The exact osmolarity for a riboflavin solution with the least shrinkage effect on biosynthetic implants needs to be found. UV-A irradiation leads to thinning of the implants in both groups (isotonic and hypotonic riboflavin solution), suggesting a direct effect of UV-A irradiation on the collagen fibrils within the implants. It is possible that the collagen fibrils became crosslinked and were hence structurally closer together, as suggested in the change in implant height (Table 1). A previous clinical phase 1 study with biosynthetic RHCIII implants showed early thinning of the implants after implantation.10 The reason was proposed to be mechanical because of the tight overlying mattress sutures. Long-term outcome 4 years after implantation demonstrated that there was no further change in corneal thickness from 1 to 4 years postoperatively, suggesting that homeostasis and deturgescence was reached.11 In the present study, thinning of the implants was induced by direct osmotic reaction and the crosslinking effect on collagen fibrils. Our results suggest that implants with a higher collagen concentration that approaches that of a normal human cornea, would likely not react as strongly to osmotic and crosslinking changes and would achieve homeostasis more rapidly. In these in vitro experiments, we only used RHCIII implants with different RHCIII concentrations. The RHCIII–MPC implants being reinforced by a second network of biopolymers, might be more stable toward mechanical und photochemical forces as proposed previously.28,29 For implantation of the biosynthetic implants in porcine and rabbit eyes, deep lamellar keratoplasty was performed by manual trephining, an excimer laser, or a femtosecond laser. The resulting corneal bed showed significant differences in its surface regularity that was dependent upon the device. Performing deep lamellar keratoplasty (DALK) with the handheld trephine or femtosecond laser produced a rather rough surface with small irregularities due to residual tissue bonds (arrowheads at the interface, Fig. 2E). The handheld trephine naturally depends on manual skills of the surgeon and can hardly be as precise as a laser system. The irregular corneal bed after performing DALK using femtosecond laser is most likely due to a problem in the docking procedure. The patient interface responsible for docking the eye to the laser is constructed for the human eye and shape of the human cornea. Porcine eyes and rabbit eyes differ even more in shape and corneal curvature, so the precision of docking process was reduced. This is probably the reason for the irregular cutting effect of the femtosecond laser. After performing DALK with the excimer laser, we could demonstrate a perfectly precise corneal bed (shown in OCT images, arrowheads at the interface, Fig. 2D; and histologically, Fig. 2F). Precise cutting leading to a smooth corneal recipient bed is elementary for accurate fitting of the implant and thus its possible fixation by riboflavin–UV-A crosslinking. Our ex vivo experiments in porcine and rabbit eyes showed that good adhesion between the biosynthetic implants
and “host” corneal bed after riboflavin–UV-A crosslinking is possible. After the crosslinking procedure, the implants remained in the corneal bed for all further examinations and demonstrated tight connection to the corneal recipient stroma in OCT images and histological sections. Because specimens that were not riboflavin–UV-A crosslinked lost the implant immediately, the crosslinking procedure must be responsible for retention of the crosslinked implants within the corneal beds. In scanning electron microscopy, direct fiber links (arrow, Fig. 2G) between the biosynthetic implant and corneal stroma were demonstrated. We assume that these fibers between implant and corneal stroma correspond to direct crosslinks that were built between collagen fibrils of the biosynthetic implant and collagen fibrils of the corneal recipient stroma and that this accounts for the adhesion of the implant. We know from numerous studies that riboflavin–UV-A crosslinking increases corneal biomechanical stability.12,13,15 The morphological correlate is assumed to be an increase in collagen fiber diameter.16 The biochemical explanation for this is that oxygen radicals produced by riboflavin–UV-A crosslinking interfere with amino acids of collagen molecules, which leads to covalent bonds in collagen fibrils.17 Riboflavin–UV-A crosslinking is widely accepted for treatment of keratoconus. Although there are few studies that report on keratoconus progression despite crosslinking,30 there are multiple studies, which have proved that riboflavin–UV-A crosslinking prevents keratoconus from progression by stabilization corneal keratometry.18–23 Although the effect of riboflavin–UV-A crosslinking is reduced with stromal depth in the human cornea, with the effect being strongest in the anterior 200 mm (65% of UV irradiation is absorbed here), the crosslinking effect also occurs between depths of 200 to 400 mm.31 Another study reported that the apoptosis of keratocytes induced by the photochemical damage of riboflavin–UV-A crosslinking was only observed in the anterior 300 mm.26 A previous study demonstrated that the cohesion between 2 stromal lamellae after riboflavin–UV-A crosslinking is not strong.32 They concluded that crosslinking stabilizes only interfibrillar and intrafibrillar, but not interlamellar, cohesion. In contrast, in the present study, we observed physical fiber links between the implant and stroma at a depth of 350 mm, at the interface between the implant and corneal stroma. We believe that these are direct crosslinks built between collagen fibrils of the implant and collagen fibrils of the corneal stroma. Our implants were cell free, comprising only recombinant human collagen and water (no proteoglycans or glycosaminoglycans in contrast to stromal lamella). We hypothesize that UV-A was therefore able to penetrate deeper unhindered and there were more exposed pendant groups in the collagen that would allow for the crosslinking effect and stronger adhesion at this depth. Furthermore, the implant and the corresponding host corneal bed shapes may have helped with the “locking-in” of the collagen fibrils from the implant to cornea. For surgical apposition, our implants were not cut perpendicular to the surface. Instead, all cuts were angled allowing for more surface area for apposition between implant and corneal bed (Fig. 2C), which may also have helped with the implant– stroma adhesion.
548
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
| www.corneajrnl.com
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
The connection between biosynthetic implant and corneal bed was more precise in the porcine eyes than in rabbit eyes. Again the cutting accuracy of the laser systems must be taken into consideration. The corneal curvature of the rabbit deviates from the human shape to a higher degree than does a porcine cornea. This accounts for a reduced adaptation of the rabbit eye to the femtosecond laser and therefore, a less precise corneal bed surface, which decreased the adhesion of the implant in the corneal bed. In our small series, we could not see a significant difference in the outcome between standard and rapid riboflavin–UV-A crosslinking. This confirms previous findings in the literature.33 Nevertheless, for use with biosynthetic implants, this needs to be confirmed in more extensive testing. In summary, we demonstrated that crosslinking with riboflavin and UV-A irradiation is a promising suture-free fixation method for biosynthetic corneal collagen implants, and a viable alternative to the use of sutures for surgical retention. Crosslinking leads to permanent adhesion of the implants in our ex vivo experiments in porcine and rabbit eyes. This could potentially reduce the suture-related complications such as haze formation, epithelial delay, and surface irregularity. The effect of riboflavin–UV-A crosslinking on the implants and their general stability, especially with regard to thinning, however, needs to be further evaluated. The long-term outcome and biocompatibility of riboflavin–UV-A crosslinking for fixation of biosynthetic corneal implants needs to be further investigated in in vivo trials. ACKNOWLEDGMENTS The authors thank Kimberley Merrett and M. Mirazul Islam, Linköping University, Sweden for making the implants and Petra Eberl for assistance with the laboratory work. REFERENCES 1. World Health Organization. Prevention of blindness and visual impairment. Available from: http://www.who.int/blindness/causes/priority/en/ index9.html. Accessed April 12, 2014. 2. Akpek EK, Alkharashi M, Hwang FS, et al. Artificial corneas versus donor corneas for repeat corneal transplantation. Cochrane Database Syst Rev. 2014;11:CD009561. 3. Magalhães FP, Sousa LB, Oliveira LA. Boston type I keratoprosthesis: review. Arq Bras Oftalmol. 2012;75:218–222. 4. Carlsson DJ, Li F, Shimmura S, et al. Bioengineered corneas: how close are we? Curr Opin Ophthalmol. 2003;14:192–197. 5. Griffith M, Osborne R, Munger R, et al. Functional human corneal equivalents constructed from cell lines. Science. 1999;286:2169–2172. 6. Li F, Carlsson D, Lohmann C, et al. Cellular and nerve regeneration within a biosynthetic extracellular matrix for corneal transplantation. Proc Natl Acad Sci U S A. 2003;100:15346–15351. 7. Liu Y, Gan L, Carlsson D, et al. A simple, cross-linked collagen tissue substitute for corneal implantation. Invest Ophthalmol Vis Sci. 2006;47: 1869–1875. 8. Merrett K, Fagerholm P, McLaughlin CR, et al. Tissue-engineered recombinant human collagen-based corneal substitutes for implantation: performance of type I versus type III collagen. Invest Ophthalmol Vis Sci. 2008;49:3887–3894. 9. Bentley E, Murphy CJ, Li F, et al. Biosynthetic corneal substitute implantation in dogs. Cornea. 2010;29:910–916.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Riboflavin–UV-A Crosslinking
10. Fagerholm P, Lagali NS, Merrett K, et al. A biosynthetic alternative to human donor tissue for inducing corneal regeneration: 24-month follow-up of a phase 1 clinical study. Sci Transl Med. 2010;2:46ra61. 11. Fagerholm P, Lagali NS, Ong JA, et al. Stable corneal regeneration four years after implantation of a cell-free recombinant human collagen scaffold. Biomaterials. 2014;35:2420–2427. 12. Spoerl E, Schreiber J, Hellmund K, et al. Untersuchungen zur Verfestigung der Hornhaut am Kaninchen [in German]. Ophthalmologe. 2000;97:203–206. 13. Wollensak G, Spoerl E, Seiler T. Stress-strain measurements of human and porcine corneas after riboflavin-ultraviolet-A-induced cross-linking. J Cataract Refract Surg. 2003;29:1780–1785. 14. Spoerl E, Wollensak G, Seiler T. Increased resistance of crosslinked cornea against enzymatic digestion. Curr Eye Res. 2004;29:35–40. 15. Wollensak G, Iomdina E. Long-term biomechanical properties of rabbit cornea after photodynamic collagen crosslinking. Acta Ophthalmol. 2009;87:48–51. 16. Wollensak G, Wilsch M, Spoerl E, et al. Collagen fiber diameter in the rabbit cornea after collagen crosslinking by riboflavin/UVA. Cornea. 2004;23:503–507. 17. Kohlhaas M. Kollagen-Crosslinking mit Riboflavin und UVA-Licht beim Keratokonus [in German]. Ophthalmologe 2008;105:785–796. 18. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-a-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol. 2003;135:620–627. 19. Hoyer A, Raiskup-Wolf F, Spoerl E, et al. Kollagenvernetzung mit Riboflavin und UVA-Licht bei Keratokonus [in German]. Ophthalmologe. 2009;106:133–140. 20. Wittig-Silva C, Chan E, Islam FM, et al. A randomized, controlled trial of corneal collagen cross-linking in progressive keratoconus: three-year results. Ophthalmology. 2014;121:812–821. 21. Caporossi A, Baiocchi S, Mazzotta C, et al. Parasurgical therapy for keratoconus by riboflavin-ultraviolet type A rays induced cross-linking of corneal collagen: preliminary refractive results in an Italian study. J Cataract Refract Surg. 2006;32:837–845. 22. Vinciguerra P, Albè E, Trazza S, et al. Refractive, topographic, tomographic, and aberrometric analysis of keratoconic eyes undergoing corneal cross-linking. Ophthalmology. 2009;116:369–378. 23. Vinciguerra P, Albè E, Trazza S, et al. Intraoperative and postoperative effects of corneal collagen cross-linking on progressive keratoconus. Arch Ophthalmol. 2009;127:1258–1265. 24. Wollensak G, Spörl E, Reber F, et al. Corneal endothelial cytotoxicity of riboflavin/UVA treatment in vitro. Ophthalmic Res. 2003;35:324–328. 25. Wollensak G, Spoerl E, Wilsch M, et al. Endothelial cell damage after riboflavin-ultraviolet-A treatment in the rabbit. J Cataract Refract Surg. 2003;29:1786–1790. 26. Spoerl E, Mrochen M, Sliney D, et al. Safety of UVA-riboflavin crosslinking of the cornea. Cornea. 2007;26:385–389. 27. Vetter JM, Brueckner S, Tubic-Grozdanis M, et al. Modulation of central corneal thickness by various riboflavin eyedrop compositions in porcine corneas. J Cataract Refract Surg. 2012;38:525–532. 28. Liu W, Deng C, McLaughlin CR, et al. Collagen-phosphorylcholine interpenetrating network hydrogels as corneal substitutes. Biomaterials. 2009;30:1551–1559. 29. Hackett JM, Lagali N, Merrett K, et al. Biosynthetic corneal implants for replacement of pathologic corneal tissue: performance in a controlled rabbit Alkali Burn Model. Invest Ophthalmol Vis Sci. 2011;52:651–657. 30. Kymionis GD, Karavitaki AE, Grentzelos MA, et al. Topography-based keratoconus progression after corneal collagen crosslinking. Cornea. 2014;33:419–421. 31. Schilde T, Kohlhaas M, Spoerl E, et al. Enzymatic evidence of the depth dependence of stiffening on riboflavin/UVA treated corneas [in German]. Ophthalmologe. 2008;105:165–169. 32. Wollensak G, Spörl E, Mazzotta C, et al. Interlamellar cohesion after corneal crosslinking using riboflavin and ultraviolet A light. Br J Ophthalmol. 2011;95:876–880. 33. Schuhmacher S, Oeftiger L, Mrochen M. Equivalence of biomechanical changes induced by rapid and standard corneal cross-linking, using riboflavin and ultraviolet Radiation. Invest Ophthalmol Vis Sci. 2011;52: 9048–9052.
www.corneajrnl.com |
549
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Riboflavin–UV-A Crosslinking for Fixation of Biosynthetic Corneal Collagen Implants Kerstin Wand, MD,* Raphael Neuhann,* Andrea Ullmann,* Katharina Plank,* Michael Baumann,† Roland Ritter,‡ May Griffith, Prof,§ Chris P. Lohmann, MD,* and Karin Kobuch, MD*
Purpose: To evaluate riboflavin–UV-A crosslinking as an alterna-
be further evaluated. Biostability, integration, and long-term outcome are further evaluated in in vivo animal experiments.
Methods: A range of cell-free corneal implants consisting of
Key Words: collagen implant, riboflavin–UV-A crosslinking, biosynthetic cornea, laser DALK, corneal transplantation, keratoprosthesis
tive suture-free fixation method for biosynthetic corneal collagen implants. recombinant human collagen type III were examined. In vitro, the implants were crosslinked with different riboflavin solutions and irradiations. Ex vivo, the biosynthetic corneal implants were placed on the anterior cornea of porcine and rabbit eyes after performing deep anterior lamellar keratoplasty with a trephine, femtosecond laser, or excimer laser. UV-A crosslinking was performed with isotonic or hypotonic riboflavin at either standard or rapid procedure. The corneas were excised, fixed in PFA 4%, and embedded in paraffin. Crosslinking effects on the implants and the adhesion between implant and corneal bed were evaluated by slit-lamp biomicroscopy, optical coherence tomography (OCT) images, and histologically.
Results: After the crosslinking procedure, the implants showed different degrees of thinning. The accuracy of cutting the corneal bed was highest with the excimer laser. Good adhesion of the implant in the corneal bed could be demonstrated in OCT images. This was more accurate in porcine eyes than in rabbit eyes. Histologically, crosslinks between implant and corneal stroma were demonstrated. There was no difference between standard and rapid crosslinking procedures.
Conclusions: Riboflavin–UV-A crosslinking as a fixation method for biosynthetic corneal collagen implants was demonstrated to be promising. It can reduce suture-related complications such as haze formation and surface irregularity. Stability of the implants, especially shrinkage after riboflavin–UV-A crosslinking, needs to Received for publication October 17, 2014; revision received January 4, 2015; accepted January 19, 2015. Published online ahead of print March 13, 2015. From the *Department of Ophthalmology, Klinikum rechts der Isar, Technische Universität München, Munich, Germany; †MLase AG, Germering München, Munich, Germany; ‡Technolas PV Bausch & Lomb, Munich, Germany; and §Department of Clinical and Experimental Medicine and Integrative Regenerative Medicine Centre, Linköping University, Linköping, Sweden. Supported by EU nanomedicine ERA-net project I-CARE, with funding from VDI (Verein Deutscher Ingenieure) and VDO to MLase AG, Swedish Research Council to MG. M. Baumann works for MLase AG; R. Ritter works for Technolas PV Bausch & Lomb. The remaining authors have no conflicts of interest to disclose. Reprints: Kerstin Wand, MD, Department of Ophthalmology, Klinikum rechts der Isar, Technische Universität München, Ismaningerstr. 22, D-81675 München, Germany (e-mail: [email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
544
| www.corneajrnl.com
(Cornea 2015;34:544–549)
C
orneal disease or damage leading to loss of corneal clarity is the ninth largest cause of permanent loss of vision or blindness worldwide.1 Currently, the most successful treatment for blindness due to corneal opacification is the transplantation of human donor corneal tissue, either by full penetrating keratoplasty or partial lamellar keratoplasty. However, grafting with donor tissue depends on the available supply of donor corneas. Because of the current shortage of acceptable donor corneas, there are long waiting lists of patients to receive a transplant. In the future, demographic age changes and an increasing number of refractive corneal surgeries will further reduce the amount of possible donor corneas. The current alternative to human donor tissues is artificial corneas or keratoprostheses, such as the Boston keratoprosthesis, the most commonly implanted artificial cornea.2 However, because of the possible severe complications and lack of host–graft integration, keratoprostheses are currently restricted to patients not suitable for donor tissue grafting, for example, in case of repeated allogeneic tissue failure or severe ocular surface diseases.3 Recently developed biosynthetic corneal implants could become a primary alternative to allogeneic transplants.4 The cornea is transparent, avascular, and immunologically privileged, which makes it an excellent candidate for tissue engineering for transplantation.5 Cell-free corneal-shaped matrices, based on crosslinked collagen, were introduced for corneal transplantation.5–8 Biosynthetic corneal implants have been applied in multiple in vivo animal experiments and showed an overall good integration in donor tissue, maintenance of optical clarity, epithelial coverage of the implants, cell and nerve ingrowth from donor tissue, and restored touch sensitivity.5–9 A phase 1 clinical study of implanted corneal substitutes (composed of EDC crosslinked recombinant human collagen type III) was conducted on 10 patients to replace the pathological anterior cornea.10 After anterior lamellar keratoplasty (ALK), the biosynthetic corneal substitutes were implanted with overlying mattress sutures that were removed after 6.5 weeks. Overall, the implants showed Cornea Volume 34, Number 5, May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
good integration in donor tissue without rejection, vascularization, infection, or the need for long-term steroid immunosuppression for more than 4 years.11 The implants were covered by stable morphologically normal epithelium, tear production and osmolarity were normal, stromal cell migration from the recipient into the implant occurred, regenerating nerves were noted after 3 months of operation, and corneal touch sensitivity returned within the first 12 months after the operation. However, suture-associated complications were noted. Areas of focal haze with thinning of the implant were found in the central cornea that corresponded to the intersection point of the overlying mattress sutures.10 The presence of tight overlying sutures had delayed epithelialization in these areas, creating an epithelial defect and subsequently an inflammatory response leading to implant thinning and invasion of repair fibroblasts. The localized thinning induced surface irregularity that was noticed postoperatively and required compensation by placement of rigid contact lenses for best-corrected visual acuity. A suture-free retention method such as riboflavin–UV-A crosslinking would therefore be expected to improve the visual outcome. Corneal crosslinking induced by riboflavin–UV-A is currently accepted for treatment for keratoconus. In numerous studies, it was shown that riboflavin–UV-A crosslinking increases long-term corneal biomechanical stability and its resistance toward collagenases.12–15 An increase in collagen fiber diameter in the anterior stroma of the cornea is assumed to be the morphological correlate to the increase of biomechanical stability.16 The biochemical background is explained by the production of oxygen radicals induced by the photosensitizer riboflavin in combination with UV-A. Oxygen radicals interfere with the amino acids of collagen molecules and this photochemical process leads to intrafibrillar and interfibrillar covalent bonds in collagen fibrils.17 In patients with keratoconus, riboflavin–UV-A corneal crosslinking has been proven to prevent the disease from progression by stabilizing corneal keratometry.18–23 It is a simple and safe procedure as long as the cytotoxic threshold of irradiance is not exceeded and the endothelium is not damaged.24–26 In this study, our purpose was to establish a new suturefree retention method for biosynthetic corneal collagen implants to resolve suture-related complications such as thinning of the implant and delay in epithelialization leading
Riboflavin–UV-A Crosslinking
to haze and surface irregularity. Using its photochemical effect of forming covalent bonds in collagen fibrils, riboflavin–UV-A corneal crosslinking could be used as an excellent tool for fixation of biosynthetic corneal implants in the recipient corneal stromal bed.
METHODS Different cell-free collagen hydrogels were used as biosynthetic corneal implants. The hydrogels were fabricated from recombinant human collagen type III (RCH III; Fibrogen Inc, San Francisco, CA) crosslinked with 1-ethyl-3-(3dimethylaminopropyl) carbodiimide and N-hydroxysuccinimide (EDC/NHS; Sigma-Aldrich, St Louis, MO). Several of these RHCIII implants were reinforced with a second network comprising 2-methacryloyloxyethyl phosphorylcholine (MPC; Biocompatibles, UK and Paramount Fine Chemicals Co Ltd, Dalian, China). Fabrication of these interpenetrating networks have been described in detail in previous publications.8,27,28 The RHCIII substitutes consisted of different RHCIII concentrations, 10, 13.7, 15, and 18%. The RHCIII–MPC substitutes consisted of RHCIII 13.7% mixed with MPC. The resulting implants were corneal shaped, 350 mm in thickness, and 12 mm in diameter. For ex vivo implantation, the corneal substitutes were either trephined to obtain 6 mm diameter buttons (Fig. 1A), or cut into shapes that complemented that of the recipient corneal bed using either an excimer laser (MLase AG, Munich, Germany) or a femtosecond laser (Technolas PV, Munich, Germany). We first ascertained that collagen hydrogels could be crosslinked with riboflavin–UV-A. RHCIII hydrogels with starting collagen concentrations of 10, 13.7, 15, and 18% were divided into 4 subgroups of 3 implants. The first group was UV-A crosslinked with isotonic riboflavin 0.1% (10 mg riboflavin-5-phosphate in 10 mL dextran T500 20% solution), the second group with hypotonic riboflavin (0.1% riboflavin in hypoosmolar solution). Riboflavin solution was applied every minute for 5 minutes before UV-A irradiation was started using the rapid protocol (18 mW/cm2 for 5 minutes) (CCL 365; MLase AG). The third and fourth groups were controls comprising implants that were treated with either isotonic or hypotonic riboflavin solution for 5 minutes only without any subsequent UV-A irradiation. Implant thickness
FIGURE 1. A, Preparation of the implant by trephining out a button of the corneal substitute. B, Corneal bed in porcine eye after deep anterior lamellar keratoplasty with excimer laser. C, Riboflavin–UV-A crosslinking of implant embedded in the corneal bed. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.corneajrnl.com |
545
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
Wand et al
was assessed before and after treatment using OCT imaging (iVue 100; Optovue Europe GmbH). Biosynthetic corneal implants (20 RHCIII hydrogels of 13.7% and 15 RHCIII–MPC hydrogels) were “implanted” into 20 ex vivo porcine eyes and 15 ex vivo rabbit eyes. The isolated globes underwent deep anterior lamellar keratoplasty ablation. Wound beds of 6 mm in diameter and 350 mm in thickness were cut into 6 porcine eyes and 250 mm in thickness into 5 rabbit eyes by handheld corneal trephine. An excimer laser (MLase AG) was used to make analog cuts in an additional 7 porcine and 5 rabbit eyes. A femtosecond laser was used on a further 7 porcine and 5 rabbit eyes (Technolas PV Bausch & Lomb) (Fig. 1B). For crosslinking, each corneal implant was positioned within its target corneal bed and riboflavin–UV-A corneal crosslinking was performed (Fig. 1C). Isotonic riboflavin 0.1% solution was applied on the contact side of the implant and stromal corneal bed for 5 minutes, allowing the riboflavin to permeate through the implant and cornea. Then, UV-A irradiation was started at either standard or rapid protocol. Directly after corneal crosslinking procedure, the implanted eyes were examined by slit-lamp biomicroscopy and OCT images (iVue 100; Optovue Europe GmbH). The corneas were then excised manually by circumferential scleral incision, fixed in paraformaldehyde 4%, and embedded in paraffin for subsequent histological examination after hematoxylin and eosin (H&E) and picrosirius red staining, as well as scanning electron microscopy (SEM). Specimens with implants treated with riboflavin but without UV-A crosslinking served as controls.
TABLE 1. Change in Thickness of RHCIII Implants After Treatment With Riboflavin Solution (Isotonic or Hypotonic), With or Without UV-A Irradiation (Rapid Protocol) Demonstrates Different Degree of Shrinkage of the Implants RHCIII 10%
13.7%
15%
18%
Average initial thickness of implant 316.41 371.41 350.75 330.90 before treatment, mm Change in thickness after treatment 253.07 236.10 228.52 237.00 with isotonic riboflavin solution, % Change in thickness after treatment 16.30 22.86 2.01 4.00 with hypotonic riboflavin solution, % Change in thickness after treatment 257.5 255.5 244.95 250.18 with isotonic riboflavin solution + UV-A irradiation, % Change in thickness after treatment +2.3 220.6 26.76 28.69 with hypotonic riboflavin solution + UV-A irradiation, %
The biosynthetic corneal implants consisting of EDC/ NHS crosslinked RHCIII of different concentrations showed different degrees of shrinkage after treatment with riboflavin solution (isotonic and hypotonic) and after riboflavin–UV-A crosslinking. Table 1 shows the average change in implant thickness (measured in micrometers in OCT images) in the 4 subgroups of each RHCIII concentration (10, 13.7, 15, and 18%) after treatment with isotonic or hypotonic riboflavin solution solely and after treatment with riboflavin solution (isotonic or hypotonic) followed by UV-A irradiation (rapid protocol, 18 mW/cm2 for 5 minutes). Table 1 demonstrates that treatment with isotonic riboflavin solution solely leads to a significant change in the RHCIII implant thickness on average by 39% compared with their initial value. Treatment with hypotonic riboflavin solution in contrast did not reduce implant thickness significantly. Treatment with isotonic riboflavin solution followed by UV-A irradiation (rapid protocol) showed an average reduction of implant thickness by 52% compared with its initial value. Treatment with hypotonic riboflavin solution followed by UV-A irradiation in contrast reduced the implant thickness on average by only 8%. Higher collagen content of RHCIII seemed to reduce the effect of riboflavin osmolarity but caused an additional shrinkage by UV-A irradiation and crosslinking (Table 1). RHCIII implants (13.7% collagen) and RHCIII–MPC implants, respectively, were implanted into ex vivo porcine and rabbit eyes. Corneal beds were prepared using 1 of the
following: a 6-mm handheld corneal trephine, an excimer laser, or a femtosecond laser. There was a noticeable difference in the surface quality of the corneal bed that was dependent on the device that was used for preparation. The excimer laser produced the smoothest surfaces with few to no irregularities. The femtosecond laser and handheld trephine produced a rather rough surface with little bumps and holes that were visible in OCT images, which led to greater difficulty in the adhesion of the implants. Riboflavin–UV-A crosslinking of the implants lying in the corneal bed resulted in permanent adhesion. The implants stayed adherent in the corneal bed for all further investigations. A scheme of the “ideal apposition” between implant and corneal bed for better understanding of the histological and OCT images is demonstrated in Figures 2A, B. Slit-lamp biomicroscopy showed tight adhesion between implant and corneal bed after the crosslinking procedure (Fig. 2C). OCT images confirmed the apposition between implant and corneal bed (Figs. 2D, E, arrowheads point at interface). After excision and fixation of the corneas, the connection between implant and corneal bed could also be demonstrated histologically (in H&E-stained and picrosirius-stained sections; Fig. 2F). In this histological section, due to implant thinning, the implant lies deep in the corneal bed and the surface of the implant and corneal recipient is not on a plane (Fig. 2F). The changes in implant thickness after riboflavin–UV-A crosslinking seen in our in vitro experiments were partially also observed in the ex vivo experiments in porcine and rabbit eyes. On higher magnification using SEM, direct fiber links between implant and corneal stroma could be found at their interface (arrow, Fig. 2G). The fitting of the implant was more accurate in porcine eyes than in rabbit eyes, and better integration could be demonstrated after deep anterior lamellar keratoplasty with excimer laser than with femtosecond laser or handheld trephine. In contrast, in the control group, in which the implants were appositioned within the surgical bed but without riboflavin–UV-A crosslinking, the implants detached completely. No further examination was possible.
546
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
RESULTS
| www.corneajrnl.com
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
Riboflavin–UV-A Crosslinking
FIGURE 2. A, Scheme of “ideal apposition” between implant and corneal bed as in OCT images. B, Scheme of “ideal adhesion” between implant and corneal stroma as in histological sections. C, Macroscopic image of RHCIII implant in a prepared corneal bed. D, Implant in porcine corneal bed after deep lamellar keratoplasty with excimer laser in OCT image (note that layered appearance of the RHCIII–MPC implant). The smooth graft–host interface is marked by arrowheads. E, Implant in rabbit corneal bed after deep lamellar keratoplasty with femtosecond laser in OCT image. The bumpy graft–host interface is marked by arrowheads. F, Good adhesion of implant to the host corneal stroma is seen in this H&E-stained section. G, Scanning electron microscopy showing physical structural connections (arrow) between implant (parallel orientation of collagen fibers) and corneal stroma.
Overall, there was no difference in the outcome between UV-A irradiation with standard (3 mW/cm2 for 30 minutes) and rapid protocols (18 mW/cm2 for 5 minutes).
DISCUSSION Biosynthetic corneal implants could become a primary alternative to transplantation of human corneal donor tissue in the treatment of corneal opacification, which would overcome the existing shortage of acceptable donor tissues.4 Multiple in vivo animal experiments and a phase 1 clinical study showed good results overall with regard to integration and optical clarity.5–11 However, there were suture-associated complications such as thinning of the implants and delayed epithelialization at earlier suture points leading to surface irregularity and haze formation.10 Therefore, a new suture-free Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
retention method would be expected to improve the visual outcome. The purpose of this study was to establish and assess an alternative suture-free fixation method using riboflavin–UVA crosslinking for incorporation of biosynthetic corneal implants in the recipient corneal stromal bed. In our in vitro experiments, the biosynthetic implants showed different degrees of shrinkage after treatment with riboflavin solution solely or riboflavin solution plus UV-A irradiation. Treatment with isotonic riboflavin solution led to significant thinning of the implants. However, treatment with hypotonic riboflavin solution did not alter the thickness, suggesting an osmotic reaction. It should be noted that the RHCIII implant comprises 85% to 90% water; so, the isotonic riboflavin solution with 20% dextran would be hyperosmolar compared with the implant, resulting in thinning. The hypotonic solution, however, is mainly water, just like the www.corneajrnl.com |
547
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Wand et al
Cornea Volume 34, Number 5, May 2015
implants and hence there was less shrinkage. Modulation of corneal thickness by various riboflavin eye drops compositions has been described previously and confirms our observation.27 The exact osmolarity for a riboflavin solution with the least shrinkage effect on biosynthetic implants needs to be found. UV-A irradiation leads to thinning of the implants in both groups (isotonic and hypotonic riboflavin solution), suggesting a direct effect of UV-A irradiation on the collagen fibrils within the implants. It is possible that the collagen fibrils became crosslinked and were hence structurally closer together, as suggested in the change in implant height (Table 1). A previous clinical phase 1 study with biosynthetic RHCIII implants showed early thinning of the implants after implantation.10 The reason was proposed to be mechanical because of the tight overlying mattress sutures. Long-term outcome 4 years after implantation demonstrated that there was no further change in corneal thickness from 1 to 4 years postoperatively, suggesting that homeostasis and deturgescence was reached.11 In the present study, thinning of the implants was induced by direct osmotic reaction and the crosslinking effect on collagen fibrils. Our results suggest that implants with a higher collagen concentration that approaches that of a normal human cornea, would likely not react as strongly to osmotic and crosslinking changes and would achieve homeostasis more rapidly. In these in vitro experiments, we only used RHCIII implants with different RHCIII concentrations. The RHCIII–MPC implants being reinforced by a second network of biopolymers, might be more stable toward mechanical und photochemical forces as proposed previously.28,29 For implantation of the biosynthetic implants in porcine and rabbit eyes, deep lamellar keratoplasty was performed by manual trephining, an excimer laser, or a femtosecond laser. The resulting corneal bed showed significant differences in its surface regularity that was dependent upon the device. Performing deep lamellar keratoplasty (DALK) with the handheld trephine or femtosecond laser produced a rather rough surface with small irregularities due to residual tissue bonds (arrowheads at the interface, Fig. 2E). The handheld trephine naturally depends on manual skills of the surgeon and can hardly be as precise as a laser system. The irregular corneal bed after performing DALK using femtosecond laser is most likely due to a problem in the docking procedure. The patient interface responsible for docking the eye to the laser is constructed for the human eye and shape of the human cornea. Porcine eyes and rabbit eyes differ even more in shape and corneal curvature, so the precision of docking process was reduced. This is probably the reason for the irregular cutting effect of the femtosecond laser. After performing DALK with the excimer laser, we could demonstrate a perfectly precise corneal bed (shown in OCT images, arrowheads at the interface, Fig. 2D; and histologically, Fig. 2F). Precise cutting leading to a smooth corneal recipient bed is elementary for accurate fitting of the implant and thus its possible fixation by riboflavin–UV-A crosslinking. Our ex vivo experiments in porcine and rabbit eyes showed that good adhesion between the biosynthetic implants
and “host” corneal bed after riboflavin–UV-A crosslinking is possible. After the crosslinking procedure, the implants remained in the corneal bed for all further examinations and demonstrated tight connection to the corneal recipient stroma in OCT images and histological sections. Because specimens that were not riboflavin–UV-A crosslinked lost the implant immediately, the crosslinking procedure must be responsible for retention of the crosslinked implants within the corneal beds. In scanning electron microscopy, direct fiber links (arrow, Fig. 2G) between the biosynthetic implant and corneal stroma were demonstrated. We assume that these fibers between implant and corneal stroma correspond to direct crosslinks that were built between collagen fibrils of the biosynthetic implant and collagen fibrils of the corneal recipient stroma and that this accounts for the adhesion of the implant. We know from numerous studies that riboflavin–UV-A crosslinking increases corneal biomechanical stability.12,13,15 The morphological correlate is assumed to be an increase in collagen fiber diameter.16 The biochemical explanation for this is that oxygen radicals produced by riboflavin–UV-A crosslinking interfere with amino acids of collagen molecules, which leads to covalent bonds in collagen fibrils.17 Riboflavin–UV-A crosslinking is widely accepted for treatment of keratoconus. Although there are few studies that report on keratoconus progression despite crosslinking,30 there are multiple studies, which have proved that riboflavin–UV-A crosslinking prevents keratoconus from progression by stabilization corneal keratometry.18–23 Although the effect of riboflavin–UV-A crosslinking is reduced with stromal depth in the human cornea, with the effect being strongest in the anterior 200 mm (65% of UV irradiation is absorbed here), the crosslinking effect also occurs between depths of 200 to 400 mm.31 Another study reported that the apoptosis of keratocytes induced by the photochemical damage of riboflavin–UV-A crosslinking was only observed in the anterior 300 mm.26 A previous study demonstrated that the cohesion between 2 stromal lamellae after riboflavin–UV-A crosslinking is not strong.32 They concluded that crosslinking stabilizes only interfibrillar and intrafibrillar, but not interlamellar, cohesion. In contrast, in the present study, we observed physical fiber links between the implant and stroma at a depth of 350 mm, at the interface between the implant and corneal stroma. We believe that these are direct crosslinks built between collagen fibrils of the implant and collagen fibrils of the corneal stroma. Our implants were cell free, comprising only recombinant human collagen and water (no proteoglycans or glycosaminoglycans in contrast to stromal lamella). We hypothesize that UV-A was therefore able to penetrate deeper unhindered and there were more exposed pendant groups in the collagen that would allow for the crosslinking effect and stronger adhesion at this depth. Furthermore, the implant and the corresponding host corneal bed shapes may have helped with the “locking-in” of the collagen fibrils from the implant to cornea. For surgical apposition, our implants were not cut perpendicular to the surface. Instead, all cuts were angled allowing for more surface area for apposition between implant and corneal bed (Fig. 2C), which may also have helped with the implant– stroma adhesion.
548
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
| www.corneajrnl.com
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Cornea Volume 34, Number 5, May 2015
The connection between biosynthetic implant and corneal bed was more precise in the porcine eyes than in rabbit eyes. Again the cutting accuracy of the laser systems must be taken into consideration. The corneal curvature of the rabbit deviates from the human shape to a higher degree than does a porcine cornea. This accounts for a reduced adaptation of the rabbit eye to the femtosecond laser and therefore, a less precise corneal bed surface, which decreased the adhesion of the implant in the corneal bed. In our small series, we could not see a significant difference in the outcome between standard and rapid riboflavin–UV-A crosslinking. This confirms previous findings in the literature.33 Nevertheless, for use with biosynthetic implants, this needs to be confirmed in more extensive testing. In summary, we demonstrated that crosslinking with riboflavin and UV-A irradiation is a promising suture-free fixation method for biosynthetic corneal collagen implants, and a viable alternative to the use of sutures for surgical retention. Crosslinking leads to permanent adhesion of the implants in our ex vivo experiments in porcine and rabbit eyes. This could potentially reduce the suture-related complications such as haze formation, epithelial delay, and surface irregularity. The effect of riboflavin–UV-A crosslinking on the implants and their general stability, especially with regard to thinning, however, needs to be further evaluated. The long-term outcome and biocompatibility of riboflavin–UV-A crosslinking for fixation of biosynthetic corneal implants needs to be further investigated in in vivo trials. ACKNOWLEDGMENTS The authors thank Kimberley Merrett and M. Mirazul Islam, Linköping University, Sweden for making the implants and Petra Eberl for assistance with the laboratory work. REFERENCES 1. World Health Organization. Prevention of blindness and visual impairment. Available from: http://www.who.int/blindness/causes/priority/en/ index9.html. Accessed April 12, 2014. 2. Akpek EK, Alkharashi M, Hwang FS, et al. Artificial corneas versus donor corneas for repeat corneal transplantation. Cochrane Database Syst Rev. 2014;11:CD009561. 3. Magalhães FP, Sousa LB, Oliveira LA. Boston type I keratoprosthesis: review. Arq Bras Oftalmol. 2012;75:218–222. 4. Carlsson DJ, Li F, Shimmura S, et al. Bioengineered corneas: how close are we? Curr Opin Ophthalmol. 2003;14:192–197. 5. Griffith M, Osborne R, Munger R, et al. Functional human corneal equivalents constructed from cell lines. Science. 1999;286:2169–2172. 6. Li F, Carlsson D, Lohmann C, et al. Cellular and nerve regeneration within a biosynthetic extracellular matrix for corneal transplantation. Proc Natl Acad Sci U S A. 2003;100:15346–15351. 7. Liu Y, Gan L, Carlsson D, et al. A simple, cross-linked collagen tissue substitute for corneal implantation. Invest Ophthalmol Vis Sci. 2006;47: 1869–1875. 8. Merrett K, Fagerholm P, McLaughlin CR, et al. Tissue-engineered recombinant human collagen-based corneal substitutes for implantation: performance of type I versus type III collagen. Invest Ophthalmol Vis Sci. 2008;49:3887–3894. 9. Bentley E, Murphy CJ, Li F, et al. Biosynthetic corneal substitute implantation in dogs. Cornea. 2010;29:910–916.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Riboflavin–UV-A Crosslinking
10. Fagerholm P, Lagali NS, Merrett K, et al. A biosynthetic alternative to human donor tissue for inducing corneal regeneration: 24-month follow-up of a phase 1 clinical study. Sci Transl Med. 2010;2:46ra61. 11. Fagerholm P, Lagali NS, Ong JA, et al. Stable corneal regeneration four years after implantation of a cell-free recombinant human collagen scaffold. Biomaterials. 2014;35:2420–2427. 12. Spoerl E, Schreiber J, Hellmund K, et al. Untersuchungen zur Verfestigung der Hornhaut am Kaninchen [in German]. Ophthalmologe. 2000;97:203–206. 13. Wollensak G, Spoerl E, Seiler T. Stress-strain measurements of human and porcine corneas after riboflavin-ultraviolet-A-induced cross-linking. J Cataract Refract Surg. 2003;29:1780–1785. 14. Spoerl E, Wollensak G, Seiler T. Increased resistance of crosslinked cornea against enzymatic digestion. Curr Eye Res. 2004;29:35–40. 15. Wollensak G, Iomdina E. Long-term biomechanical properties of rabbit cornea after photodynamic collagen crosslinking. Acta Ophthalmol. 2009;87:48–51. 16. Wollensak G, Wilsch M, Spoerl E, et al. Collagen fiber diameter in the rabbit cornea after collagen crosslinking by riboflavin/UVA. Cornea. 2004;23:503–507. 17. Kohlhaas M. Kollagen-Crosslinking mit Riboflavin und UVA-Licht beim Keratokonus [in German]. Ophthalmologe 2008;105:785–796. 18. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-a-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol. 2003;135:620–627. 19. Hoyer A, Raiskup-Wolf F, Spoerl E, et al. Kollagenvernetzung mit Riboflavin und UVA-Licht bei Keratokonus [in German]. Ophthalmologe. 2009;106:133–140. 20. Wittig-Silva C, Chan E, Islam FM, et al. A randomized, controlled trial of corneal collagen cross-linking in progressive keratoconus: three-year results. Ophthalmology. 2014;121:812–821. 21. Caporossi A, Baiocchi S, Mazzotta C, et al. Parasurgical therapy for keratoconus by riboflavin-ultraviolet type A rays induced cross-linking of corneal collagen: preliminary refractive results in an Italian study. J Cataract Refract Surg. 2006;32:837–845. 22. Vinciguerra P, Albè E, Trazza S, et al. Refractive, topographic, tomographic, and aberrometric analysis of keratoconic eyes undergoing corneal cross-linking. Ophthalmology. 2009;116:369–378. 23. Vinciguerra P, Albè E, Trazza S, et al. Intraoperative and postoperative effects of corneal collagen cross-linking on progressive keratoconus. Arch Ophthalmol. 2009;127:1258–1265. 24. Wollensak G, Spörl E, Reber F, et al. Corneal endothelial cytotoxicity of riboflavin/UVA treatment in vitro. Ophthalmic Res. 2003;35:324–328. 25. Wollensak G, Spoerl E, Wilsch M, et al. Endothelial cell damage after riboflavin-ultraviolet-A treatment in the rabbit. J Cataract Refract Surg. 2003;29:1786–1790. 26. Spoerl E, Mrochen M, Sliney D, et al. Safety of UVA-riboflavin crosslinking of the cornea. Cornea. 2007;26:385–389. 27. Vetter JM, Brueckner S, Tubic-Grozdanis M, et al. Modulation of central corneal thickness by various riboflavin eyedrop compositions in porcine corneas. J Cataract Refract Surg. 2012;38:525–532. 28. Liu W, Deng C, McLaughlin CR, et al. Collagen-phosphorylcholine interpenetrating network hydrogels as corneal substitutes. Biomaterials. 2009;30:1551–1559. 29. Hackett JM, Lagali N, Merrett K, et al. Biosynthetic corneal implants for replacement of pathologic corneal tissue: performance in a controlled rabbit Alkali Burn Model. Invest Ophthalmol Vis Sci. 2011;52:651–657. 30. Kymionis GD, Karavitaki AE, Grentzelos MA, et al. Topography-based keratoconus progression after corneal collagen crosslinking. Cornea. 2014;33:419–421. 31. Schilde T, Kohlhaas M, Spoerl E, et al. Enzymatic evidence of the depth dependence of stiffening on riboflavin/UVA treated corneas [in German]. Ophthalmologe. 2008;105:165–169. 32. Wollensak G, Spörl E, Mazzotta C, et al. Interlamellar cohesion after corneal crosslinking using riboflavin and ultraviolet A light. Br J Ophthalmol. 2011;95:876–880. 33. Schuhmacher S, Oeftiger L, Mrochen M. Equivalence of biomechanical changes induced by rapid and standard corneal cross-linking, using riboflavin and ultraviolet Radiation. Invest Ophthalmol Vis Sci. 2011;52: 9048–9052.
www.corneajrnl.com |
549
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
