Therapeutic Keratoplasty

Key Concepts

Therapeutic keratoplasty is indicated for recalcitrant infectious, inflammatory, or noninflammatory conditions of the cornea, including trauma, which threaten the integrity of the globe.
Penetrating and lamellar are the two types of therapeutic keratoplasty, and each has advantages and disadvantages that must be considered when performing this type of corneal transplant.
Advantages of anterior lamellar keratoplasty compared to penetrating keratoplasty include absence of endothelial rejection, reduced topical steroids, and associated glaucoma; and maintenance of globe integrity during surgery with lower risk of iris or lens extrusion, suprachoroidal hemorrhage, endophthalmitis, or postoperative globe rupture.
Preoperative and postoperative care involve maximally reducing infection/inflammation with medical therapy while maintaining globe integrity.
Outcomes of therapeutic keratoplasty have a less favorable prognosis compared to optical keratoplasty, but good outcomes and comparable results between penetrating and lamellar keratoplasty have been demonstrated.

Therapeutic keratoplasty is a surgical procedure whose primary purpose is either to restore the structural integrity of the eye (tectonic keratoplasty) or to resolve an infectious or inflammatory keratitis that is refractory to conventional medical therapy. Often, both of these indications may be present. Therapeutic keratoplasty is generally an emergent or semiemergent procedure in which the survival of the globe is at stake. Unlike optical keratoplasty, where visual restoration is of primary importance, visual outcomes of therapeutic keratoplasty are of secondary importance to eradicating infection and preserving the structural integrity of the globe. If needed, visual rehabilitation can always be accomplished at a later date under more controlled circumstances.

The two established methods for therapeutic keratoplasty are penetrating keratoplasty and lamellar keratoplasty. Penetrating keratoplasty involves removing a full-thickness button of cornea and replacing it with a corneal graft, whereas a lamellar therapeutic keratoplasty involves replacing only the anterior cornea with a partial-thickness corneal graft, leaving the posterior cornea intact. In the presence of corneal perforation or deep stromal infection, penetrating keratoplasty is usually the preferred option.

In recent years there has been some shift toward therapeutic lamellar keratoplasty. If corneal endothelium and Descemet membrane are unaffected by the disease process, replacement of the anterior cornea may restore vision without some of the risks associated with penetrating keratoplasty. With anterior lamellar keratoplasty, there is no risk of endothelial rejection, decreased need for postoperative steroids, and lower rate of endothelial cell loss. Preserving Descemet membrane reduces the risk of intraoperative suprachoroidal hemorrhage, iris or lens extrusion, postoperative endophthalmitis, or ruptured globe. Excellent therapeutic and visual results are achievable with lamellar keratoplasty, especially with advances in deep lamellar dissection such as Anwar’s big bubble, femtosecond laser, or other techniques that allow cleavage within 50 μm of Descemet membrane. ^, ^, Vision may be compromised by optical interference at the graft–host interface caused by disparities in the donor tissue and the recipient bed, especially when the stroma is manually dissected anterior of Descemet membrane. Varying forces in the graft or remaining stromal bed may create striae in the posterior cornea. Stromal rejection occurs in up to 25% of anterior lamellar keratoplasties but readily responds to treatment with topical steroids. ^,

Therapeutic Keratoplasty for Infectious Keratitis

Despite continuing advances in the medical management of infectious keratitis, it continues to be a major cause of global blindness, often affecting marginalized and indigent populations. Proper isolation and identification of the causative organism remains a critical step in the treatment algorithm. While identification of the organism can prove useful, there remains a subgroup of bacteria, fungi, amoeba, parasites, and viruses that does not respond to antimicrobial therapy. Therapeutic keratoplasty is indicated when inflammatory or infectious corneal disease progresses despite maximal medical therapy, and the integrity of the globe is at risk. Medical therapy should be continued or altered, if necessary, until there is no reasonable expectation that the infection or inflammation can be controlled or until there is a significant risk of corneal perforation or scleral extension. Corneal perforation dramatically increases the risk of endophthalmitis and reduces the survival of the keratoplasty, and progression to scleral involvement should be avoided owing to the severe morbidity of scleritis. Less invasive surgical treatment, such as conjunctival flaps and tarsorrhaphies, can be attempted if indicated. Recent studies have also demonstrated the potential of corneal collagen crosslinking using riboflavin and UVA irradiation in treating medically refractory infectious keratitis to avoid surgical intervention.

Therapeutic keratoplasty provides a surgical debridement of an infectious process. The goal of therapeutic keratoplasty is to remove the infectious inoculum completely or to decrease the organisms in the cornea to a level at which exogenous antiinfective/antiinflammatory agents and the patient’s endogenous host defense mechanisms can be effective. ^,

In general, therapeutic keratoplasties constitute a minority of all keratoplasties performed worldwide. However, a review of 875 penetrating keratoplasties in China by Wang et al. reported infectious keratitis to be the leading indication (37.1%) for the procedure. Al-Yousuf et al. reviewed 784 penetrating keratoplasties performed in the UK, of which 8.3% were for therapeutic reasons including unresponsive infection, threatened perforation, and actual perforation. More recently, Zare et al. reviewed 1859 corneal transplants from Iran, 10.1% of which were indicated due to infectious corneal ulcers.

The causative organisms responsible for cases of infectious keratitis that need therapeutic keratoplasty also vary geographically, but bacterial infections tend to predominate. Out of 80 consecutive therapeutic keratoplasties reviewed by Killingsworth et al., 26 were performed for bacterial infections, 15 for fungal infections, 11 for active herpetic keratitis, 2 for Acanthamoeba keratitis, and the remaining 26 for noninfectious indications. A review in India of 100 therapeutic keratoplasties by Sony and colleagues also demonstrated that the majority of the procedures were performed for refractory bacterial keratitis. In contrast, recent studies from India and East Asia demonstrated a predominance of therapeutic keratoplasties for fungal keratitis over bacterial keratitis. Sharma et al. published a case series of 506 eyes that underwent therapeutic keratoplasty for infectious keratitis. Bacterial infections predominated (31.0%), followed by fungal (20.9%), mixed bacterial/fungal (6.9%), viral (5.3%), and Acanthamoeba (1.6%). Coagulase-negative Staphylococcus was the most common bacteria, and Aspergillus was the most common fungus. Follow-up ranged from 10 to 42 months and showed success in the restoration of tectonic integrity in 89.7% of eyes.

When an acute ulceration threatens perforation, consider systemic as well as topical therapy to reduce the likelihood of intraocular involvement. Subconjunctival antibacterial agents do not appear to offer a significant advantage over aggressive topical therapy. Subconjunctival and intrastromal antifungal agents have been shown in few case series to increase success in treating refractory fungal keratitis. Once an acute infectious corneal perforation has occurred, topical antimicrobials may be continued, with consideration of the potential risk of corneal and retinal toxicity. Closing the perforation with cyanoacrylate adhesives restores the integrity of the anterior chamber, opens the angle, and prevents the formation of synechial angle-closure glaucoma. While this method may, in some instances, be considered definitive treatment, bacterial infections have been shown to progress beneath the cyanoacrylate glue, so the patient must be followed closely. It is important to continue antimicrobial therapy to optimize preoperative conditions if therapeutic keratoplasty becomes necessary.

When perforations or descemetoceles are treated with therapeutic penetrating keratoplasty, outcomes are generally satisfactory. Jonas et al. examined a series of 60 cases of keratoplasty for these indications and found that 90% achieved some improvement in visual acuity. Ten patients (17%) required repeat keratoplasty for recurrent corneal ulcers. In a series of 20 patients who experienced a corneal perforation graft, the rate of transparency was 67% in 15 eyes that underwent central penetrating keratoplasty, and 85% of eyes had improved visual acuity following their therapeutic keratoplasty.

Therapeutic lamellar keratoplasty may also be effective in the treatment of refractory infectious keratitis. In a study of 126 eyes, Anshu documented equivalent success in graft survival and infection eradication outcomes comparing therapeutic deep anterior lamellar keratoplasty to therapeutic penetrating keratoplasty, and significantly better visual outcomes in the lamellar keratoplasty group. However, important to note is that lamellar keratoplasties were reserved for cases of bacterial, fungal, or acanthamoeba keratitis without perforation or intraocular involvement.

Therapeutic Keratoplasty for Bacterial Infections

Most bacterial infections, when managed expediently, respond to therapy, but due to delay in patients seeking or obtaining medical treatment, the infections may rapidly progress to corneal perforation. Certain bacteria, notably Pseudomonas aeruginosa , can infect the cornea and progress to corneal perforation over 24–48 hours. These bacteria may produce collagenase, which results in rapid corneal thinning. Refractory or acutely severe bacterial keratitis has been treated with therapeutic keratoplasty in an attempt to reduce the infectious load in the cornea as well as to prevent and/or treat perforation. In combination with medical therapy, the procedure can provide a cure rate for bacterial keratitis of up to 100%. ^, One study showed success in 147 of 157 (93.6%) cases of isolated bacterial keratitis treated with therapeutic keratoplasty. Satisfactory 1-year outcomes in therapeutic keratoplasty for advanced suppurative bacterial keratitis have also been demonstrated.

Advanced bacterial keratitis may require a larger area of corneal removal and graft placement. Although larger graft size has been associated with decreased graft survival, Cowden et al. showed that using grafts 9.5 mm or larger ( Fig. 141.1 ) can salvage an infected eye that may otherwise need to be enucleated.

Fig. 141.1, Therapeutic keratoplasty for bacterial infection.

Infectious crystalline keratopathy is a rare subgroup of refractory keratitis most commonly caused by α-hemolytic Streptococcus , but can be seen with a variety of bacteria as well as Candida . These infections are most commonly seen following corneal surgery (e.g., penetrating or lamellar keratoplasty, limbal relaxing incisions, laser in situ keratomileusis [LASIK]) with prolonged corticosteroid therapy, and may slowly but relentlessly progress. In cases unresponsive to prolonged topical fortified vancomycin or other medical management, therapeutic keratoplasty is generally successful since there is minimal inflammation in this disorder.

The role and success of therapeutic lamellar keratoplasty is increasing in the treatment of bacterial keratitis. In a 1971 study of Pseudomonas ulcers treated with therapeutic keratoplasty, Malik and Singh reported that all eight lamellar grafts became reinfected when compared to 4 of the 26 penetrating grafts. In 2007, Ti et al. reported success in 12 therapeutic lamellar keratoplasties for acute infectious keratitis with only one failed graft caused by recurrence of infection. Diagnostic acumen and medical treatment coupled with patient selection have yielded improved results over time with this treatment.

Therapeutic Keratoplasty for Fungal Keratitis

Fungal ulcers account for a relatively small percentage of total infectious keratitis cases, but in tropical climates and developing countries can account for up to 50% of infectious corneal ulcers. ^, In warmer climates filamentous fungi such as Fusarium and Aspergillus predominate, whereas in temperate climates yeasts such as Candida are more common. In the United States contact lens wear has been identified as a risk factor, and previous outbreaks have been attributed to contact lens solution. ^,

Fungal ulcers often have worse outcomes than bacterial ulcers. Fungal keratitis may respond to appropriate topical antifungal agents such as natamycin, amphotericin B, or voriconazole, depending on species and extent of involvement. The Mycotic Ulcer Treatment Trial (MUTT) 1 study demonstrated significantly better outcomes with topical natamycin than topical voriconazole in the treatment of filamentous fungal ulcers. The voriconazole group was associated with higher rates of corneal perforation and therapeutic keratoplasty. The MUTT 2 study demonstrated no additional benefit of adding oral voriconazole to treatment of filamentous fungal keratitis. Secondary analysis of MUTT data suggests that Fusarium keratitis might benefit from treatment with oral voriconazole in addition to topical natamycin. Subconjunctival and intrastromal antifungal injection have showed varied success in treating refractory cases, with randomized controlled studies needed to determine their benefit. ^,

With incomplete evidence-based literature to guide treatment, therapeutic keratoplasty remains an important and viable option for the appropriate patient. Forster and Rebell emphasized the importance of intensive antifungal treatment prior to therapeutic keratoplasty to help stabilize the cornea and improve the prognosis. They reviewed 61 eyes with corneal fungal infections, of which 13 were considered treatment failures. Nine of the 13 patients with treatment failure underwent therapeutic keratoplasty, with five of these nine patients achieving a visual acuity of 20/70 or better. In all nine patients who underwent penetrating keratoplasty, the fungal disease process was halted by the therapeutic keratoplasty. Forster observed that 17 of 29 corneas infected by fungi remained culture positive at the time of keratoplasty, and 26 of the eyes had viable hyphal elements on pathology. Postoperatively, fungal infection may infiltrate the grafted tissue.

Despite this, Killingsworth et al. obtained a 100% cure rate in 15 fungal ulcers treated with therapeutic penetrating keratoplasty. Xie and coauthors studied 108 eyes with fungal keratitis in which therapeutic penetrating keratoplasty was performed. Eighty percent of eyes remained clear during follow-up with no recurrence of infection and visual acuity ranged from 20/100 to 20/20. A later study by Xie and colleagues investigated the effectiveness of lamellar keratoplasty in treating fungal keratitis. They achieved therapeutically beneficial results in 93% of eyes (51 of 55 operations performed), with a resulting visual acuity ranging from 20/63 to 20/20. In four cases, there was a recurrence of the fungal infection within 2 weeks, which was cured by a therapeutic keratoplasty. In 2006, in a large Chinese study of 604 therapeutic keratoplasties for fungal keratitis, the authors demonstrated preservation of globe integrity and useful visual acuity in 95.7% of the eyes. Fungal infection recurred in 14 (3.5%) of the 399 penetrating keratoplasties and 13 (7.3%) of the 177 lamellar keratoplasties. Corneal tissue used in lamellar keratoplasty may be obtained more readily than healthy tissue used in penetrating keratoplasty, further advocating lamellar keratoplasty as a viable option for treating fungal keratitis.

The mainstay for treatment after therapeutic keratoplasty has been corticosteroids to prevent allograft rejection and control postoperative inflammation. However, in the management of mycotic keratitis, corticosteroids have been discouraged because of the risks of exacerbating an existing infection or introducing a superinfection. Corticosteroid use has also been implicated as a risk factor for fungal recurrence after surgical intervention. In a prospective, nonrandomized interventional case series of three patients, topical cyclosporine A (tCSA) 0.5% has been shown to be a safe and useful adjunct in the treatment of therapeutic keratoplasty for fungal keratitis to avoid or reduce the use of corticosteroids. Furthermore, compared to corticosteroids, tCSA has been shown in vitro to have a statistically significant suppressive effect on fungal growth.

Therapeutic Keratoplasty for Acanthamoeba Keratitis

The respective roles of medical and surgical intervention in Acanthamoeba keratitis are controversial. Many patients diagnosed early have been successfully treated with topical medication alone, with single or combination therapy with polyhexamethylene biguanide (0.02%), and chlorhexidine (0.02%) or propamidine (0.1%). ^, The acute management of active cases is to sterilize the infection as rapidly as possible and to delay surgical management until the patient receives adequate antiamoebic therapy. A review of 116 patients medically treated for Acanthamoeba keratitis showed infection resolution in 66.7% of eyes with pharmacotherapy alone and 33.3% of eyes needing penetrating keratoplasty. In a 1993 report by Ficker et al., graft survival for Acanthamoeba keratitis was poor, with more than 50% incidence of recurrence in the graft. Better outcomes were shown by Robaei et al. in 2015, with only seven of the 26 therapeutic penetrating keratoplasty patients requiring repeat keratoplasties and only one of those caused by recurrence of infection. In a seven-patient experience, Cullen et al. concluded that early diagnosis helped with successful medical treatment, but penetrating keratoplasty continues to have a central role in the management of cases that are advanced or unresponsive to medical therapy. Cryotherapy and crosslinking of the host cornea for cases unresponsive to medical treatment have achieved limited and variable results.

Therapeutic lamellar keratoplasty has been used to treat Acanthamoeba keratitis successfully. Nine cases of Acanthamoeba keratitis treated with lamellar keratoplasty were reviewed by Anshu et al. and only one of the infections recurred postoperatively. They concluded that lamellar keratoplasty provides a viable alternative to penetrating keratoplasty in progressive microbial keratitis.

In a series of 32 patients, the most common complication of therapeutic keratoplasty for Acanthamoeba keratitis was glaucoma, with more than half of these patients developing graft failure and mydriatic, fixed pupils.

Therapeutic Keratoplasty for Viral Keratitis

With a chronic, indolent, and recurrent nature, the course of viral keratitis is less predictable than bacterial or fungal keratitis. Affecting 1.5 million people globally, herpes simplex keratitis is the most common cause of unilateral infectious corneal blindness in the developed world, with less common causes including varicella-zoster virus, cytomegalovirus, and Epstein-Barr virus. Be wary that acanthamoeba and, to a lesser extent, fungal keratitis, are often misdiagnosed as herpetic disease.

Herpetic viral disease of the cornea may require surgical intervention. Most commonly, this is done for significant corneal scarring and is performed under quiescent, controlled circumstances. Adjunctive medical management has made possible successful penetrating keratoplasty in the active stromal stage of herpes keratitis, although operating on a quiet eye without active keratitis or uveitis is preferred. Therapeutic keratoplasty can be employed for significant ulceration and/or perforation, or to remove viral antigenic material responsible for repeated immune inflammatory episodes ( Fig. 141.2 ).

Fig. 141.2, Therapeutic keratoplasty for herpetic keratitis.

Oral acyclovir, valacyclovir, or ganciclovir are likely as effective as topical antivirals for the treatment of herpetic epithelial keratitis, avoid ocular surface toxicity, and may benefit treatment of stromal or intraocular herpetic disease. Oral or topical antiviral prophylaxis during the immediate postoperative period, and during allograft rejection episodes while topical corticosteroids are used, may prevent recurrent herpetic disease. Prolonged prophylactic use of oral antivirals reduces the incidence of herpes simplex recurrence by 45%. A large multicenter randomized trial is underway to determine if prolonged valacyclovir treatment may reduce complications or recurrence of herpes zoster eye disease (NCT03134196).

Ficker et al. found that the combination of prophylactic antivirals and corticosteroids increased the success rate in inflamed eyes undergoing therapeutic keratoplasty for herpes simplex keratitis to a rate comparable to that of quiescent eyes. The use of interrupted sutures, prompt removal of loose sutures, and antiviral and topical steroid treatment combination reduced the recurrence of herpes simplex virus to 15%. Killingsworth et al. divided 15 patients who required therapeutic penetrating keratoplasty for herpes simplex keratitis into two groups. The first group included patients with severe stromal keratitis who did not respond to medical treatment and progressed to perforation. The second group consisted of patients who developed corneal perforation secondary to persistent epithelial defects with little or no active stromal inflammation. Only 4 of the 11 patients in the first group achieved clear grafts, while all four patients in the second group were successful. Three additional studies documented recurrence rates of 9%–46%, ^, ^, with similar rates of visual rehabilitation and morbidity obtained with penetrating keratoplasty or deep lamellar keratoplasty.

The prognosis for therapeutic keratoplasty in herpes zoster keratitis is generally worse than for herpes simplex keratitis, often caused by ocular surface compromise. Patients with herpetic corneal disease usually have neurotrophic corneas; however, the anesthesia tends to be more severe for herpes zoster than for herpes simplex. One of the most important prognostic variables for herpes zoster therapeutic keratoplasty is the level of corneal sensation. Therapeutic keratoplasty in patients with herpes zoster keratitis perforations often requires adjunctive therapy, such as ocular surface lubrication, punctal occlusion, tarsorrhaphy, or conjunctival flap. Aggressive management of external and corneal disease can improve the outcomes in therapeutic keratoplasty for varicella-zoster keratitis. Tanure et al. published results from a series of 15 keratoplasties performed for varicella-zoster keratopathy from 1989 through 1998 and found that 87% of grafts remained clear at an average follow-up of 50 months. To reduce complications from neurotrophic keratopathy, four eyes received lateral tarsorrhaphies in conjunction with the keratoplasties, and frequent lubrication was prescribed.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here