A, A fourth-generation dendrimercomposed of glycerol and succinic acid. B, A first-generation dendritic polymer,derivatized with methacrylate moieties, composed of polyethylene glycol, succinicacid, and glycerol.
A, Incision using a 4.1-mm keratomeknife. B, Application of the adhesive. C, Application of the argon ion laservia a handheld probe.
A, Appearance of the globe afterthe application of glue and laser. B, Appearance of the globe after sutureplacement.
Arrangement of the experimentusing a cardiac transducer to monitor the intraocular pressure of the repairedeyes. BSS indicates balanced salt solution.
Velazquez AJ, Carnahan MA, Kristinsson J, Stinnett S, Grinstaff MW, Kim T. New Dendritic Adhesives for Sutureless Ophthalmic Surgical ProceduresIn Vitro Studies of Corneal Laceration Repair. Arch Ophthalmol. 2004;122(6):867-870. doi:10.1001/archopht.122.6.867
To compare a biodendrimer adhesive with a conventional suture for repairinglinear and stellate corneal lacerations.
A keratome knife was used to create 4.1-mm full-thickness linear incisions(n = 36) and 3 × 4-mm full-thickness stellate incisions (n = 25) inthe central cornea of enucleated human eyes. The incisions were sealed witheither a suture or the biodendrimer adhesive. The globes were inflated withbalanced salt solution, and the increase in intraocular pressure was monitoredvia a cardiac transducer until fluid leaked from each eye. This intraocularpressure reading from the transducer was recorded at the sight of any leakagethrough the wound (leakage pressure). By using the Wilcoxon rank sum test,the median leakage pressure was compared for each closure method, separatelyfor each wound group. By using the 1-sided 1-sample t test,each mean leakage pressure value was compared with 34 mm Hg, which is theintraocular pressure under certain stressful physiologic conditions (eg, coughingand the Valsalva maneuver).
For globes that underwent a linear incision, the mean leakage pressurewas 78.7 mm Hg for the sutured group and 109.6 mm Hg for the adhesive group.Globes that underwent a stellate incision had a mean leakage pressure of 57.8mm Hg for the sutured group and 78.7 mm Hg for the adhesive group. All ofthese pressures showed a statistical significance from 34 mm Hg via a 1-sided1-sample t test.
The difference in leakage pressures for all 4 groups was statisticallysignificant relative to 34 mm Hg. This suggests that either method of closure,adhesive or suture, can withstand physiologic increases in intraocular pressurepostoperatively and that biodendrimer adhesives are able to seal large corneallacerations.
The use of biodendrimer adhesives to repair a corneal wound constitutesa viable alternative clinical procedure to conventional sutures.
In human patients, corneal perforations are repaired with sutures. However,the use of sutures has limitations and drawbacks. First, suture placementitself inflicts trauma to corneal tissues, especially when multiple passesare necessary. Second, sutures such as 10-0 nylon are not without problems.Suture material may incite infection, inflammation, and vascularization; cornealscarring is more prone to develop with inflammation and vascularization present.1 Third, because of uneven tension on the sutures, asymmetrichealing and irregular astigmatism may arise.2 Thepostoperative integrity of the sutures may also be problematic, with sutureloosening or breaking requiring timely removal. Last, a prolonged operativetime and technical skill are needed for effective suture placement.
Tissue adhesives are an attractive alternative to sutures. Cyanoacrylateadhesives were first used in the 1960s by Webster et al3 inthe repair of corneal perforations. These adhesives are an effective therapeuticoption in certain ophthalmic settings. These settings include emergency treatmentof small corneal perforations and prophylactic treatment of progressive cornealthinning disorders.4 The goal of tissue adhesivetherapy frequently is to provide immediate restoration of structural integrityand occasionally to prevent further corneal thinning. In either case, cornealadhesives can lead to permanent corneal healing or at least offer temporaryclosure for the anticipation of further surgical intervention that may benecessary (eg, corneal patch grafting, lamellar keratoplasty, and therapeuticpenetrating keratoplasty).5
However, corneal adhesives such as cyanoacrylate also have limitationswith regard to their ease of applicability and effectiveness.6- 9 Themethods of application vary and can be cumbersome. The technique requiresadept and delicate application of a precise amount of adhesive in a dry environmentto facilitate wound closure. Therapeutic effectiveness is often restrictedto small corneal perforations of usually less than 1 to 2 mm because of theinability to close larger perforations. Cyanoacrylates polymerize quicklyinto a hard and brittle material. Patients can experience discomfort fromthis material in the eye and, thus, a bandage contact lens is often used.Cyanoacrylate adhesives also become opaque when they polymerize, thus obscuringthe view of the underlying structures. Furthermore, complications with cornealcyanoacrylate adhesives include cataract formation, corneal infiltrations,granulomatous keratitis, glaucoma, and even a toxic reaction in the retina.10- 13
Recently, a photo–cross-linkable tissue adhesive composed of hyaluronicacid was reported for the repair of corneal lacerations. This adhesive formulationwas applied to 38 experimental corneal linear and stellate wounds (size, 3mm) in rabbits and subsequently irradiated with a low-intensity argon laserbeam to produce a clear flexible polysaccharide hydrogel patch that sealedthe wound. The corneal perforations were completely sealed and the anteriorchambers had reformed by 6 hours in hyaluronic acid adhesive–treatedeyes. There was no evidence of leakage at this or later times in 37 of the38 eyes. The intraocular pressure (IOP) had increased to near normal levelsby day 7 in all treated eyes. The clinical examination results showed minimalinflammation, and this observation was consistent with the histological findings.14 However, this adhesive is best for repairing full-thicknesswounds, and is not effective for other ophthalmic surgical procedures.
Dendrimers offer innovative solutions to address tissue-engineeringchallenges such as the design of novel tissue adhesives. Unlike linear polymers,like cyanoacrylate or hyaluronic acid, dendrimers are composed of a specificnumber of branched repeat units that emanate from a central core (Figure 1A and B).15- 17 Thehighly branched framework provides many peripheral functional groups thatcan be modified based on the desired application. In fact, through the properselection of core and branched repeat composition and number of branched repeatunits, dendrimers can be tailored for specific applications.18- 22
We are interested in the design of biodendrimers (dendrimers that arecomposed of natural metabolites or materials known to be biocompatible) formedical applications.23,24 Recently,the synthesis of zero- to fourth-generation dendritic polymers composed ofpolyethylene glycol, succinic acid, and glycerol was reported.25 Thesedendrimers were derivatized with methacrylate moieties and tested as cornealadhesives. Based on mechanical properties and photo–cross-linking rates,the ([G1]-PGLSA-MA)2-PEG dendritic macromolecule (Figure 1B) was identified as a viable candidate for use as a cornealadhesive.
This study evaluates this new argon ion laser–activated biodendrimeradhesive for the repair of corneal lacerations in enucleated human eyes.
Each enucleated human eye was held under the operating microscope sothat the cornea was oriented upwards, facing the microscope. A 4.1-mm metalkeratome knife (Alcon Laboratories, Inc, Ft Worth, Tex) was then used to makea full-thickness 4.1-mm linear incision in the central cornea in 36 eyes (Figure 2A). Of these eyes, 27 were repairedwith our biodendrimer adhesive and 9 were repaired with 10-0 monofilamentnylon (Ethicon, Inc, Somerville, NJ). A 3 × 4-mm full-thickness stellateincision was made in an additional 25 eyes. Of these eyes, 12 were repairedusing the biodendrimer adhesive and 13 were repaired using 10-0 nylon. Allwounds were positive for the Seidel sign and were not self-sealing incisions.
For the linear and stellate incision adhesive groups, approximately15 to 20 µL of the biodendrimer was applied to the dried wound usinga 30-gauge needle (Becton Dickinson & Co, Rutherford, NJ) attached toa 1-mL syringe (Becton Dickinson & Co). Attention was given to ensurethat the adhesive was applied to the inner borders of the laceration and thesurface of the corneal laceration. Dried wounds resulted in better controlof the biodendrimer, whereas a wet surface resulted in dilution of the biodendrimerand difficult polymerization. By using a handheld probe, the argon ion laser(a diffuse beam; 200 mW; pulse duration, 1 second) (Coherent, Inc, Santa Clara,Calif) was used to polymerize the biodendrimer and form a semitransparentcross-linked gel, closing the wound (Figure2B and C). A photoinitiator dye present in the biodendrimer solutionwas used to start the cross-linking reaction and to monitor the course ofthe photolysis reaction. After approximately 20 to 30 seconds, the solutionturned from pink to clear, indicating that the reaction was complete and thatthe adhesive had formed a gel; therefore, polymerization was complete (Figure 3A). For the sutured group, 3 interrupted10-0 nylon sutures were used to close the 4.1-mm linear incision using a needleholder and 0.12 forceps (Figure 3B).The stellate incisions were repaired using 4 interrupted 10-0 nylon sutures(1 suture on each branch of the incision).
A cardiac transducer (Hewlett Packard, Palo Alto, Calif) was used tomonitor the IOP of the repaired eyes, as done in similar experiments describedin the literature.26 A cardiac transducer wasprimed and a 20-gauge needle (Sherwood Medical, St Louis, Mo) was attachedto the end of the tubing leading from the bottle of balanced salt solution(Figure 4). The needle was theninserted into the optic nerve approximately 1 cm into the globe. The needleremained in the globe, and no movement was noted. It was not necessary totie the needle and optic nerve together to secure the needle. The cardiacmonitor was placed on an arterial pressure setting and adjusted to 0 mm Hg.A 30-gauge needle on a balanced salt solution–filled 5-mL syringe (BectonDickinson & Co) was inserted through the sclera at the pars plana (ie,approximately 4 mm from the limbus). Balanced salt solution was slowly injectedinto the eye via a syringe pump (Hewlett Packard) to slowly increase the IOPas measured by the transducer. A handheld IOP-measuring device (Tono-Pen;Medtronic Solan, Jacksonville, Fla) was used to confirm the concordance ofthe transducer readings. Surgical eye spear sponges (Opticel; Wilson OphthalmicCorp, Mustang, Okla) were used to wipe the laceration site to check for signsof leakage through the corneal wound. The IOP reading from the transducerwas recorded when leakage through the wound (the leakage pressure) was observed.
Linear sutured wounds were compared with linear wounds that receivedthe biodendrimer adhesive. The mean, standard deviation, and median leakagepressures were compared for the 2 groups. The P valuewas calculated using the Wilcoxon rank sum test to determine whether the leakagepressures were statistically different between the 2 groups. This was similarlydone for the globes that underwent a stellate incision.
The linear wounds that received sutures (n = 9) had a mean leakage pressureof 78.7 mm Hg (SD, 27.8 mm Hg). The median leakage pressure was 76.0 mm Hg.The minimum and maximum values were 20 and 117 mm Hg, respectively. Examinationof the linear wounds that received the adhesive (n = 27) revealed a mean leakagepressure of 109.6 mm Hg (SD, 82.7 mm Hg). The median leakage pressure was96.0 mm Hg. The minimum and maximum values were 16 and 360 mm Hg, respectively.The calculated P value for the median leakage pressurebetween these 2 groups was .39 by the Wilcoxon rank sum test.
The stellate wounds that received sutures (n = 13) had a mean leakagepressure of 57.8 mm Hg (SD, 28.9 mm Hg). The median leakage pressure was 51.0mm Hg. The minimum and maximum values were 25 and 125 mm Hg, respectively.Analysis of the stellate wounds that received a polymer (n = 12) showed amean leakage pressure of 78.7 mm Hg (SD, 59.6 mm Hg). The median value was68.5 mm Hg. The minimum and maximum leakage pressures were 10 and 220 mm Hg,respectively. By using the Wilcoxon rank sum test, P =.43 was calculated using the median leakage pressure between these 2 groups.
By using a 1-sided 1-sample t test, we comparedthe mean leakage pressure for all 4 experimental groups with 34 mm Hg. Wechose this IOP because it has been reported that the IOP increases up to 60%under physiologic functions such as coughing, the Valsalva maneuver, and exercising.27 There was a statistically significant differencein the mean leakage pressure for all 4 groups compared with 34 mm Hg: forstellate wounds that received an adhesive, P = .01;for stellate wounds that received a suture, P = .006;for linear wounds that received an adhesive, P<.001;and for linear wounds that received a suture, P =.001.
The mean leakage pressure was higher for the stellate and linear woundsthat received the biodendrimer sealant compared with a suture. However, thisdifference in mean leakage pressure for both types of wounds was not statisticallysignificant when the sutured corneas were compared with the corneas that receivedthe adhesive. The linear and stellate wounds that received a polymer alsohad a higher standard deviation compared with globes that received sutures.Whether the incision was sutured or sealed, the wounds were secure. The repairedwounds were sufficiently strong to withstand leakage pressures measured inreference to an IOP of 34 mm Hg, which corresponds to normal physiologic stressessuch as coughing, the Valsalva maneuver, and exercising.
With each use of the biodendrimer adhesive, our technique improves andwe observe better outcomes. This adhesive is more user friendly and requiresless technical expertise. The globes that received the dendritic adhesivehad a higher mean leakage pressure, although this was not statistically significant.In addition, this sealant is elastic, which will reduce the likelihood forastigmatism, and is degradable because it possesses ester linkages, suggestingthat cell migration into the wound can occur during healing. Other advantagesof this biodendrimer adhesive over suture or cyanoacrylate adhesives includeits control of polymerization, its transparency and elasticity, and its smoothrubbery texture when gelled. Possible future applications of our polymer mayinclude its use in securing corneal transplants and dislocated laser-assistedin situ keratomileusis flaps or its use as a drug delivery vehicle. We areconducting in vivo studies on white leghorn chickens (Gallusdomestica) with similarly constructed linear wounds that are eithersealed with the photo–cross-linkable biodendrimer adhesive or sutured.In summary, this sealant holds promise as a replacement or supplement to standardsutures used in the repair of small and large corneal lacerations.
Corresponding author and reprints: Terry Kim, MD, Duke UniversityEye Center, Campus Box 3802, Erwin Road, Durham, NC 27710 (e-mail: firstname.lastname@example.org).
Submitted for publication July 9, 2003; final revision received October22, 2003; accepted November 21, 2003.
This study was supported by grant R01 EY13881-01 from the National EyeInstitute, National Institutes of Health, Bethesda, Md; the Pew Foundation,Philadelphia, Pa; and the Johnson & Johnson Focused Giving Program, NewBrunswick, NJ.
Dr Velazquez had full access to all the data in the study and takesresponsibility for the integrity of the data and the accuracy of the dataanalysis.