Article Text

Download PDFPDF

Surgical adhesives in ophthalmology: history and current trends
  1. Samantha Guhan1,
  2. Si-Liang Peng2,3,
  3. Hrag Janbatian1,2,3,
  4. Stephanie Saadeh2,
  5. Stephen Greenstein3,
  6. Faisal Al Bahrani2,4,
  7. Ali Fadlallah1,2,3,5,
  8. Tsai-Chu Yeh2,6,
  9. Samir A Melki1,2,3
  1. 1 Harvard Medical School, Boston, Massachusetts, USA
  2. 2 Boston Eye Group, Boston, Massachusetts, USA
  3. 3 Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston, Massachusetts, USA
  4. 4 Boston University, Boston, Massachusetts, USA
  5. 5 Faculty of Medicine, Saint Joseph University, Beirut, Lebanon
  6. 6 University of California, San Francisco, San Francisco, California, USA
  1. Correspondence to Dr Samir A Melki, Cornea and Refractive Surgery Service, Massachusetts Eye & Ear Infirmary, Boston, MA 02114, USA; samir_melki{at}


Tissue adhesives are gaining popularity in ophthalmology, as they could potentially reduce the complications associated with current surgical methods. An ideal tissue adhesive should have superior tensile strength, be non-toxic and anti-inflammatory, improve efficiency and be cost-effective. Both synthetic and biological glues are available. The primary synthetic glues include cyanoacrylate and the recently introduced polyethylene glycol (PEG) derivatives, while most biological glues are composed of fibrin. Cyanoacrylate has a high tensile strength, but rapidly polymerises upon contact with any fluid and has been associated with histotoxicity. Fibrin induces less toxic and inflammatory reactions, and its polymerisation time can be controlled. Tensile strength studies have shown that fibrin is not as strong as cyanoacrylate. While more research is needed, PEG variants currently appear to have the most promise. These glues are non-toxic, strong and time-effective. Through MEDLINE and internet searches, this paper presents a systematic review of the current applications of surgical adhesives to corneal, glaucoma, retinal, cataract and strabismus surgeries. Our review suggests that surgical adhesives have promise to reduce problems in current ophthalmic surgical procedures.

  • treatment surgery
  • wound healing
  • cornea
  • retina
  • glaucoma
View Full Text

Statistics from


Tissue adhesives could potentially improve outcomes of current ophthalmic surgical procedures, reduce surgical time and minimise complications such as postoperative leaks, inflammation and postoperative infections. An ideal tissue adhesive should have superior tensile strength, be non-toxic and anti-inflammatory, improve efficiency and be cost-effective. Adhesives are available in two primary categories: synthetic glues (including cyanoacrylate and polyethylene glycol (PEG) derivatives) and biological glues (such as fibrin), as shown in table 1. The goal of this paper is to provide a comprehensive literature review using MEDLINE and internet searches of the current applications of surgical adhesives in corneal, glaucoma, retinal, cataract and strabismus surgeries.

Table 1

Commercially available cyanoacrylate, fibrin and polyethylene glycol (PEG)-based surgical adhesives

Cyanoacrylate glue is formed from an exothermic reaction between formaldehyde and cyanoacetates. Cyanoacrylate forms a very strong seal, but also rapidly polymerises upon contact with any fluid and has been associated with histotoxicity.1 Higher molecular weight cyanoacrylates (n-butyl (Histoacryl, TissueSeal, Ann Arbor, Michigan, USA)), iso-butyl and octyl-cyanoacrylates (Dermabond, Ethicon, Somerville, New Jersey, USA) are stronger and less toxic than their lower molecular weight equivalents (methyl-cyanoacrylate).2 3

Fibrin adhesives are made using two components that are mixed together to form the final insoluble glue. The first component of fibrin products like Tisseel (Baxter International, Deerfield, Illinois, USA) and Tissucol (Baxter Hyland Immuno, Vienna, Austria) includes fibrinogen, human factor XIII and aprotinin. The second component contains thrombin and calcium. When mixed, thrombin breaks the fibrinogen into fibrin monomers, which are then polymerised into a linear fashion and crosslinked by calcium-activated human factor XIII. Aprotinin prevents breakdown of the fibrin glue. Different fibrin products have slight variations in their components. For example, Quixil (OMRIX Biopharmaceuticals, New York, New York, USA) uses tranexamic acid to prevent fibrinolysis, while Evicel (OMRIX Biopharmaceuticals, New York, New York, USA) includes neither aprotinin nor tranexamic acid. Fibrin is biocompatible and biodegradable, and thus induces less toxic and inflammatory reactions. Furthermore, its polymerisation time can be controlled, as the reaction rate is determined by the concentration of thrombin.4 5 However, tensile strength studies have shown that fibrin is not as strong as cyanoacrylate.6 The Evicel variant is supposed to be the strongest fibrin version yet and could potentially improve adhesion strength.4 7 One disadvantage of the fibrin glues is the risk of infectious disease transmission since the adhesive is created from blood donation.8 This can be mitigated by using autologous fibrin.

PEG sealants are formed by crosslinking PEG polyethers with target tissue proteins.9 These glues are non-toxic and relatively strong. Polymerisation can be induced in some models by xenon light, while other variants such as ReSure (PEG with trilysine acetate, Ocular Therapeutix, Bedford, Massachusetts, USA) allow a 30 s window for the adhesive to be applied before it polymerises.10 11

Surgical application techniques

Cyanoacrylates polymerise upon contact with physiological fluids. Since they have a relatively low viscosity, Bhatia recommends that patients lie supine during ophthalmic procedures to prevent run-off.12 Polymerisation times vary from product to product. Mixture with iophendylate can be used to delay polymerisation.13 In our experience, it is best to be conservative with the initial cyanoacrylate glue application as the material will expand and may lead to significant irritation. One can apply a drop of glue on the operating tray and merely touch it with the tip of a 24 gauge needle or catheter, thereby harvesting a thin layer of glue by capillarity. This is then applied to the area of interest.14 Some physicians also choose to apply the cyanoacrylate with a sponge soaked in the adhesive.15 When applied to the cornea, it is important to make sure the epithelium is denuded and dried in the areas surrounding the lesion as the glue will not adhere to corneal epithelium.14

Fibrin adhesive preparation involves creating the two-component mixtures according to manufacturer’s instructions. Application technique varies from surgery to surgery. Tisseel (Baxter International) can be applied with Baxter’s recommended double-barrelled injection system, but many researchers do not use this system because the amount injected is too large and there is no control over the mixing time of the components.12 Instead some applied fibrin using 2 mL syringes with 26 gauge needles and applied 1 drop of each component simultaneously.16 Bhatia recommended 1 cm3 syringes with 27 gauge needles.12 Banitt et al sealed corneal incisions by applying each component separately to stromal edges.17 Hovanesian and Karageozian also recommended dispensing each component separately.18 Polymerisation time can be delayed if needed by diluting the thrombin concentration with balance saline solution. The 1:10 and 1:100 dilutions delay polymerisation by 1 and 2 min, respectively.18 Evicel fibrin (OMRIX Biopharmaceuticals, Somerville, New Jersey, USA) comes with a spray applicator, but that should only be used if one can accurately predict the spray distance.19 Tachosil fibrin (Baxter Biosurgery, Deerfield, Illinois, USA) is a surgical patch that can be cut to a desired size and placed in a desired location.20

Different PEG commercial products vary significantly in their application. CoSeal (Baxter International) is applied with a double-barrel injector. Bhatia recommends that the patient lies supine because the glue runs. Since the glue expands to four times its size, less glue is recommended. CoSeal takes about 60 s to polyermise.12 OcuSeal (Beaver-Visitec International, Waltham, Massachusetts, USA) is applied using a brush and takes about 20 s to polymerise.21ReSure (Ocular Therapeutix) is also mixed according to the manufacturer’s instructions and allows a 20 s window before polymerisation. ReSure is currently Food and Drug Administration (FDA) approved to seal clear corneal incisions (CCIs) up to 3.5 mm in cataract surgery.22 FocalSeal (Genzyme Corporation, Cambridge, Massachusetts, USA) polymerises within 40–60 s after illumination by a xenon arc lamp (450–500 nm, blue-green).11

Pterygium surgery

Fibrin glue has been used in pterygium surgery for over 30 years and has been proven to safely attach amniotic membrane and conjunctival grafts during wound closure. Numerous studies suggested that compared with suturing, fibrin is helpful in reducing surgical time, improving postoperative discomfort, avoiding suture-associated complications and reducing reoccurence.23

A meta-analysis published in 2017 analysed whether fibrin glue was more efficacious than sutures at binding conjunctival autografts during pterygium surgery. It showed that the fibrin experimental group had a lower recurrence rate compared with suture group, at >6 months follow-up, with no difference in complication rates.24 The healing in the fibrin group had little inflammation, less discomfort and better aesthetic outcome.24 Based on the article written by Suzuki et al, nylon and silk sutures tend to cause an immune reaction and Langerhans cell migration to the cornea.25 It is believed that this inflammatory response is what favours recurrence.

Glued autologous conjunctival grafting is the preferred method in reducing recurrence rates and complications. Various techniques have evolved from the traditional method developed by Kenyon et al in 1985.26 The conventional method using fibrin (Tisseel, Baxter International) was to mix the fibrinogen and thrombin components before its application onto the scleral bed and then positioning the conjunctival autograft. Koranyi et al described the cut and paste technique where the two components of the glue were used separately. Fibrinogen was placed on the graft, epithelium side down and thrombin was placed on the sclera. The graft had to be quickly flipped onto the bare sclera and smoothed out in 30 s to be attached firmly.27 In both cases, the main drawback was the rapid clot formation after graft positioning. The sandwich fibrin glue technique published in 2013 by Fava et al allowed more time for graft manipulation prior to clot formation. This method created a fibrinogen–conjunctival autograft–thrombin sandwich; the fibrinogen is placed first on the scleral bed, followed by the autograft epithelium side up and finally the thrombin component, once the graft is smoothed and in perfect position.28

Limbal stem cell transplantation

Both cyanoacrylate glue and fibrin glue have been widely used in limbal stem cell transplantation. Many studies have shown positive results by using cyanoacrylate glue during the process of dissection and harvestation of limbal stem cell allografts from corneoscleral buttons.29 The cyanoacrylate is used to anchor the donor corneoscleral rim to a block, providing stability that aids the lamellar dissection by minimising damage to the limbal epithelium. In addition, fibrin glue has fixated keratolimbal allografts and living-related conjunctival limbal allograft in people with stem cell deficiency.30 Suturing may lead to postoperative infection, inflammation, vascularisation and provoke graft rejection.30 31 Thus, fibrin sealants have become a rational alternative that reduces suture-related complications and improves surgical time and patient comfort.23 31


Managing leaks after glaucoma surgery is critical to preventing the formation of peripheral synechiae, cataracts, severe hypotony, hypotony-induced maculopathy, endophthalmitis and choroidal effusion.32 Synthetic and biological glues have been used in glaucoma drainage device surgeries, especially to block postoperative leaks from the drainage tube’s insertion site or from a bleb.

Bahar et al studied human fibrin glue, Quixil (OMRIX Biopharmaceuticals, Somerville, New Jersey, USA), in trabeculectomy.33 The scleral flap and conjunctiva were sealed with fibrin adhesive in eight eyes, while six control eyes had sutures. Both groups had resolution of minimal congestion and chemosis within 10 days, no corneal haze, no gross anterior chamber reaction and no reported positive Seidel test. An active filtering bleb presented in both groups, except for one experimental eye. While both groups had significantly decreased intraocular pressures (IOPs) after 1 day postoperatively, Bahar and his colleagues attribute the resurgence in pressure to the fast scar formation in rabbit models.33

Bahar et al also studied Quixil glue in five adults whose varying forms of glaucoma were not responsive to drugs.34 Each had elevated IOPs, ‘visual acuity worse than hand movement’, eye pain due to the pressures and total optic disc cup. The mean preoperative pressure was 40 mm Hg. The scleral and conjunctival flaps were sealed in place with Quixil glue. Afterwards, all had active, filtering blebs with no leakage, and three were medication-free, as their pressures were between 10 and 20 mm Hg after 1 year. One patient’s pressure stabilised through topical treatment, whereas the other patient’s pressure remained elevated. With no postoperative complications, Bahar and his colleagues concluded the general outcome was positive.34

Kusabara and Kasahara applied synthetic cyanoacrylate glue to a leaky insertion site of a glaucoma drainage device.35 Following device implant surgery, a patient had a flat chamber, low IOP and choroidal folds. N-butyl-2-cyanoacrylate was applied to the leaking area around the tube, which was enclosed with a patch graft from cadaver sclera. After several weeks, the anterior chamber was deep and without leakage, although the authors do not comment on the final visual acuity.

Valimaki used Tisseel fibrin glue after insertion of a Molteno valve implant.36 All Molteno tubes were closed with a 7–0 Rapid Vicryl suture (which absorbs in about 2 weeks), and if a leak around the scleral flap area was observed in an eye, Tisseel was applied along its edges. From 11 patients, 11 eyes demonstrated intraoperative leaking, so fibrin glue was applied over the scleral flap intraoperatively, with a mean, preoperative IOP of about 35 mm Hg. After application of Tisseel in 11 eyes in a 3-week postoperative period, the pressure for nine eyes dropped down to less than 22 mm Hg and stabilised at this range. The other two eyes experienced reduced pressures but ‘failed’, as the parameters for a successful pressure reduction were greater than or equal to 6 mm Hg and less than or equal to 22 mm Hg.36

Asrani and Wilensky compared biological fibrin glue with other sealants for sealing bleb leaks, after performing a retrospective chart review in a 12-month period.37 Initial treatment of 35 bleb leaks involved patching, followed by (if needed): autologous fibrin tissue glue, cyanoacrylate glue, suturing, bandage contact lens, collagen shield and injection of autologous blood into the bleb. Of the 35 leaks, patching treated six. Autologous fibrin tissue glue stopped leakage in 9 out of 12 episodes. Cyanoacrylate glue treated three of eight patients. Suturing was successful in 4 of 15 instances. With spontaneous resolution seen in two, bandage contact lens, collagen shield and injection of blood under the conjunctiva did not resolve their respective leaks. Asrani and Wilensky concluded autologous fibrin tissue glue may be the best among these non-surgical treatment modalities for bleb leaks.37

Refractive surgery

Epithelial ingrowth treatment

Fibrin tissue adhesives have gained more ground in refractive surgery for treatment of recalcitrant cases of epithelial ingrowth since the invention of laser in situ keratomileusis (LASIK). The incidence of epithelial ingrowth in microkeratome-assisted LASIK ranges reportedly from 0% to 20% but only 1.7% are clinically significant and require surgical removal.38–40 With the use of femtosecond laser-assisted LASIK, the rate of epithelial ingrowth has decreased significantly. Kamburoglu and Ertan noted an incidence of 0.03% in their series, with a recurrence rate of 1.8%.40 The reason for the higher incidence in microkeratome-assisted LASIK is hypothesised to be related to either epithelial cell implantation during surgery dragged by the microkeratome, or fluid carrying floating epithelial cells under the flap, or epithelial cells invading under the edge of the flap creating a fistula connecting the host surface epithelium.40–42

Epithelial ingrowth necessitates treatment when it enters the visual axis, affects visual acuity and induces astigmatism. It also prompts intervention when it causes glare, foreign body sensation, flap irregularity and corneal melt. The suggested method of management is debridement and possibly flap suturing to create a tight barrier between the stromal bed and the flap to prevent recurrent ingrowth. Anderson and Hardten used Tisseel glue in the LASIK wound edge and debridement with adjunctive alcohol to treat three patients with recurrent epithelial ingrowth.43 Narvaez et al demonstrated that debridement, flap suturing and fibrin glue application can eliminate extreme epithelial ingrowth.44 In 2010, Rapuano treated a significant recurrence using the similar method described by Narvaez and found no subsequent recurrence.45

The main disadvantage is that fibrin glue becomes opaque when polymerised, thus making it difficult to visualise interface inflammation. It usually takes approximately 1–2 weeks for the spontaneous dissolution of the glue. There are no randomised studies so far proving the benefit of glue as superior to the other approaches to preventing recurrent epithelial ingrowth, such as suturing alone.

Corneal flap tears and dislocations

In 2005, Yeh et al published a case using fibrin glue and mechanical debridement to treat traumatic LASIK flap dislocation and epithelial ingrowth. The patient had bilateral LASIK 21 months prior to flap dislocation. After treatment, no subsequent recurrent epithelial ingrowth was noted after 20 months follow-up and vision return to normal.46

Corneal surgery

The use of glue in corneal surgery has a long track record when it comes to preventing impending perforations. Significant efforts are now spent to eliminate or replace sutures during corneal transplantation. The latter will greatly enhance the efficiency of and recovery from anterior corneal transplantation.

Corneal perforations and melts

In the 1960s, Refojo et al and Webster et al pioneered the use of tissue adhesives in the treatment of corneal perforations.16 47 Since then, research studies have shown that cyanoacrylate effectively treats corneal perforations up to 3.0 mm in diameter.16 48 49 Corneal patching along with tissue adhesives is used as a temporising measure reducing enucleation rates.49

Cyanoacrylate glue can arrest the progression of corneal necrosis by inhibiting polymorphonuclear leucocytes with strong collagenolytic and proteolytic activities.49 50 In 1983, Eiferman and Snyder concluded that the glue inhibits growth of Gram-positive organisms.49 51 This type of adhesive has been used in various types of corneal perforations ranging from infectious to traumatic. The main concern for cyanoacrylate derivatives is that direct contact is toxic to the lens and corneal endothelium. This may lead to iridocorneal or iridolenticular adhesions.49 51 Some studies have noted that cyanoacrylate glues can induce a foreign body reaction and corneal neovascularisation.

Sharma et al found that both fibrin and N-butyl-2-cyanoacrylate are effective in treating corneal perforations up to 3 mm in diameter.16 However, fibrin glue leads to faster healing (63–31.8% in 6 weeks) and less corneal vascularisation compared with cyanoacrylate glue.16 The main difference between the two glues is that fibrin has a longer plug formation time.16

For perforations or melts greater than 3 mm, the use of amniotic membrane or scleral patch graft augmented with cyanoacrylate or fibrin glue can be used for temporary closure.50 Grau and Duran treated a 3 mm corneal perforation by sandwiching the perforation between a double-layered amniotic membrane and a single-layered amniotic membrane and using TachoSil applied at the perforation site to glue the membranes together.52

Corneal transplants

From as early as the 1970s, Rosenthal et al attached lamellar corneal grafts on rabbits using a platelet/fibrinogen/thrombin mixture.53 In 2003, Kaufman et al used human fibrin (Tisseel) for sutureless lamellar keratoplasty in five patients. All five lamellar grafts, measuring 200 μm in thickness, healed and remained clear postoperatively and had a final visual acuity of 20/40—20/20.54 In 2007, Duarte and Kim  used the sandwich technique for lamellar keratoplasty, similar to the one proposed for pterygium surgery, where the fibrinogen is applied first on the stromal bed followed by the thrombin on the donor button. This allows more time for graft manipulation and allows for more control over coagulation case initiation. The final visual acuity in Duarte’s case report was 20/30.55

In the same year, Narendran et al performed deep anterior lamellar keratoplasty using overlay sutures and fibrin adhesive. They found the technique to be successful when both the donor button and the recipient bed were the same size and thickness. Their preferred method of using the fibrin glue (Tisseel) is to split the two preprepared solutions into two different syringes to allow better control of application.56 Once the corneal button was in place, a small amount of each solution was applied thinly on the recipient bed. The overlay sutures were removed after 1 month postoperatively and the bandage contact lens was left in situ for 1 week after suture removal. The graft remained clear at 6 months after all steroid drops were stopped at 4 months postoperatively, with a final visual acuity of 6/12.56

Tissue adhesives were also used in full thickness corneal transplantations. After the pioneering study done by Katzin using fibrin glue for penetrating corneal grafts in rabbits in the 1940s, tissue adhesives have gained new ground with the advent of femtosecond lasers.57 In recent years, femtosecond-laser-assisted keratoplasty has raised in popularity due to the pretended potential of faster visual recovery, better wound stability and apposition.58 One of the main advantages of femtosecond laser over traditional trephination is that the laser can be programmed to cut the exact shape, whether it be ‘top hat’, ‘zig-zag’, ‘mushroom’ or ‘Christmas tree’ configuration onto the graft and donor, making a pretended more precise fit. In 2007, Bahar et al used a femtosecond laser to evaluate the stability of different wound arrangements. This paper demonstrated that the use of fibrin glue in Top Hat femtosecond-laser-assisted penetrating keratoplasty (PK) was more mechanically stable and had less astigmatism compared with standard PK, PK using other configurations or Top Hat PK with sutures. In the Top Hat PK with fibrin group, wound leakage occurred at >158 mm Hg after placement of 16 sutures compared with 103.8 mm Hg and 59.1 mm Hg in the Top Hat PK without glue and traditional PK groups, respectively.59

Recently, fibrin glue has even been found useful in preparing grafts used in Descemet stripping automated endothelial keratoplasty (DSAEK). Sharma et al described a novel technique by using Tisseel to fixate surrogate scleral rims in donor corneas with inadequate scleral rim for successful microkeratome-assisted dissection. This method can expand the donor tissue pool for DSAEK.60

Cataract surgery

Cataract surgery employs CCIs or scleral incisions for phacoemulsification and IOL implantation. The wound is then either sutured or left to self-seal. Currently, a vast majority of surgeons leave the CCI to self-heal, leading to the possibility of wound leakage. Wound leakage could occur whenever there is a variation in IOP, such as when a patient rubs their eye or blinks. Transient, decreased IOP post-surgery could also lead to fluid and bacterial ingress into the eye.61 A study performed by Wallin et al demonstrated that wound leak on the first postoperative day increased the risk of bacterial endophthalmitis by 44-fold in cases in which intracameral antibiotics were not used.62 Thus, investigation of surgical adhesives to close CCIs during cataract surgery seems beneficial.

In 1996, Alio et al successfully used Histoacryl cyanoacrylate to close 6.5 mm-wide scleral tunnel incisions, which led to a lower occurrence of astigmatism in the 12-week follow-up period compared with sutures.63 In 2005, Meskin et al, Ritterband et al and Leung et al independently studied the use of 2-octyl cyanoacrylate in humans.64–66 While the adhesive closed CCIs in all cases, Meskin et al noted that some patients complained of a mild foreign body sensation and diffuse bulbar conjunctival hyperaemia.64 In 2009, Banitt et al compared the efficacy of N-butyl-2-cyanoacrylate, Tisseel fibrin and sutures to seal 3 mm, 4.5 mm and 6 mm incisions by measuring the IOP required to make the sealed wound leak (IOPL). They found that N-butyl-2-cyanoacrylate adhesive required a statistically significant higher IOPL for the 4.5 mm incision, but not for the 3 mm or 6 mm incision as compared with fibrin and sutures.17 In 2012, Kaja et al demonstrated that 2.5 mm incisions sealed with N-butyl cyanoacrylate were able to withstand much higher IOPs than wounds sealed with sutures.67

In 1998, Henrick et al used Tisseel fibrin to seal scleral pocket incisions without the need of additional sealants or sutures.68 While Mester et al demonstrated that the use of Tissucol fibrin (Baxter Hyland Immuno) reduced surgically induced astigmatism, Alio et al did not find any significant difference.69 70 In 2007, Hovanesian and Karageozian found that Tisseel fibrin glue both prevented ingress with variations in IOP and significantly raised the amount of pressure needed for fluid egress.18 In the 2009 Banitt et al comparative study, researchers found that Tisseel fibrin required a statistically significant higher IOPL than sutures in sealing 3 mm incisions, but not in sealing 4.5 mm or 6 mm incisions.17

PEG adhesives have become one of the most popular sealants for cataract surgery. Maddula et al found that OcuSeal (powder PEG and liquid polyethylene amine; Beaver-Visitec International) raised the IOP needed for fluid egress.71 Uy and Kenyon found that OcuSeal led to less surgically induced astigmatism and foreign body sensations than sutures. Compared with the control group, OcuSeal resulted in significantly less foreign body sensation, but no significant difference in surgically induced astigmatism.21 Calladine et al, Dell et al, and Masket et al showed that ReSure sealant also resulted in a higher IOP, stopped micro-leaks and led to less wound leakage than self-sealed CCIs.10 72 73

Retina surgery

Current methods for retinal surgery have multiple downfalls that could be rectified by surgical adhesives. Vitrectomy requires a patient to lie in a specific orientation for weeks post-surgery, while tamponade cannot fix injuries at particular locations in the retina and can lead to complications like glaucoma and cataracts.74

In 1978, Faulborn and Witschel successfully used Histoacryl cyanoacrylate to glue the retina to the choroid after an open-sky vitrectomy in two patients.75 In the late 1980s, McCuen et al reported the successful use of a transvitreal Histoacryl cyanoacrylate retinopexy in rabbits and humans.76 77 An additional study demonstrated that the transvitreal cyanoacrylate retinopexy performed with silicone oil lead to a ‘crescent shaped tear at the edge of the glue site with subsequent retinal detachment’, and thus did not recommend the use of silicone oil.78 Spitznas et al emphasised that only small amounts of Histoacryl can be used in order to avoid toxicity.79 80

Fibrin glues have also had success in retinal surgeries. Tilanus and Deutman showed that Tissucol successfully closed macular holes and Coleman et al demonstrated that Tisseel led to reattachment of small retinal tears.81 82 However, Coleman et al also found that the glue was unable to seal large retinal tears and stressed that the tissue glue alone was not enough to ensure permanent retinal reattachment.82 Al Sabti et al used Tisseel fibrin sealant to seal optic disc pit-associated macular detachment in two out of three patients for at least 3 years.83 This required patients to stay in prone position for 2 weeks postoperatively. In 2017, de Oliveira used the Evicel Fibrin as an adjunct in optic disc pit-associated macular detachment and concluded that Evicel did not result in a significant inflammatory reaction and reduced subretinal fluid in all four patients.7 Also in 2017, Ozdek and Ozdemir used autologous fibrin to seal persisting macular holes and found that the retinas remained attached for at least 14 months.84

Hoshi et al demonstrated that FocalSeal PEG (Genzyme Corporation) successfully sealed in vitro porcine retinal tears.11 Hoshi et al also used PEG to seal sclerotomies post vitrectomy surgery in rabbit eyes.85 No significant inflammation or toxicity was observed in either study. Sarfare et al found Medicus PEG (Medicus Bioscience, San Jose, California, USA) did not lead to any inflammatory or toxic response 2 weeks post injection into the subretinal space of albino mice.86 Hubschman found that Medicus PEG led to retinal attachment at least 8 weeks post surgery; an inflammatory reaction was also observed in 4 out of the 12 eyes.87

Strabismus surgery

Investigating adhesive use in strabismus surgery began many years ago. Saving time, money and elimination of the risk of retinal tears are the main motives driving continued research today. Inadequate bond strength remains the top concern.

One of the first studies to quantitatively measure bond strength was published in 1969 by Dunlap et al.88 It assessed bond strength and biocompatibility in rabbits. Bond strengths were followed for 5 days and histological changes were followed for 4 months. The study showed that the average weight required to dislocate a glue bond was <66 g  compared with <238 g with sutures 72 hours postoperatively. Histological studies showed that increased thickness of the glue layer might cause inadequate bond strength. The muscle was found to extend forward beyond the glue in an attempt to reach the sclera, eventually jeopardising strength and precision.

Adal-1, a synthetic mixture of ethyl cyanoacrylate and ethyl carboxyacrylate was studied by Mulet et al in 2006 for use in strabismus surgery in humans. The control group consisted of 17 eyes undergoing recession with sutures (7/0 vicryl).89 The study group had 10 eyes where the muscle and sclera were attached via Adal-1. Muscle displacement and inflammation were evaluated for 3 months postoperatively. Both groups showed no muscle breaks from insertion site. While displacement in the control group and Adal-1 group did not significantly differ (P=0.459), the control group did significantly experience more immediate postoperative conjunctival irritation (P=0.040).

In 2009, a study was performed in rabbits to evaluate the use of a new long chain cyanoacrylate tissue adhesive.90 It provided tensile strength similar to older generations but with an inflammatory reaction rate closer to fibrin. This was the first study looking at use of adhesives in resection surgery. This study by de Alba Campomanes et al compared 2-octyl-cyanoacrylate with 6–0 polyglactin 910 suture after primary recession and reoperation for advancement of the superior rectus muscle. Twenty-seven rabbits were randomised. One eye was operated with cyanoacrylate and the other with polyglactin suture. Cyanoacrylate was considered non-inferior to suture for both the operation and the reoperation procedures. A statistical difference was found between the mean slippage after advancement compared with that found after the first recession procedure for both techniques.90

Bona and Arthur tested a technique that incorporated a polyglactin mesh attached to the sclera using cyanoacrylate. This technique might prevent unwanted retinal tears due to sutures but it is not meant to save time as the muscle should later be approximated to the sclera using a suture that passes only through the mesh.91 The study was conducted on cadaver eyes. The results showed linear increase in tensile strength with both increased time given to polymerise and increased surface area. The bond had a tensile force in excess of the maximum force generated by extraocular muscles in humans.92

Fibrin is the most studied biological sealant in strabismus surgery. Its use for extraocular muscle surgeries has been tested previously and proven to have low tensile strength. Yang et al compiled a review report to assess the use of fibrin for conjunctival wound closure after strabismus surgeries.93 This paper studied five level II evidence papers and showed that fibrin reduces postoperative inflammation and patient discomfort compared with sutures. But combined results from three of the five papers studied in this review showed that 2 of 50 eyes (total from three papers) (4%) required sutures to repair a retracted conjunctival wound that was initially closed using fibrin glue. Time needed to close conjunctival wounds was less with fibrin compared with sutures but skilled surgeons might complete the procedure in a similar time frame.

In conclusion, biological glues like fibrin lack the tensile strength to be used in extraocular muscle surgeries, but can be incorporated in conjunctival wound closure if post-op irritation from suture is a concern. Synthetic glues like Adal-1 show promising results and can be possibly incorporated in recession surgeries to reduce operative time.


This review demonstrates the major efforts spent through the past few years on using glue in ocular surgery. There is convincing evidence of the glue benefits in many areas. Further research should focus on improving chemical composition, adherence strength and methods of application.


View Abstract


  • Contributors SAM had the original idea for the article and is the guarantor. All other authors were involved in literature search, drafting, revising and final approval of the document.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement There is no additional unpublished data from this study.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.