Corneal Graft Rejection 

Updated: Oct 16, 2018
Author: Michael Taravella, MD; Chief Editor: Hampton Roy, Sr, MD 



Although described for more than 100 years, corneal transplantation has become increasingly common since the 1960s. In 2016, approximately 83,000 corneal transplantations were performed in the United States.[1] Corneal graft rejection is the most common cause of graft failure in the late postoperative period.[2]

Examples of corneal graft rejection are shown in the images below.

This severely vascularized cornea would be at high This severely vascularized cornea would be at high risk for graft rejection following a penetrating keratoplasty. This patient experienced Stevens-Johnson syndrome.
This is an example of an acute graft rejection epi This is an example of an acute graft rejection episode. Note the graft edema, Descemet folds, and keratic precipitates.


Corneal transplantation has a high success rate in part because of the relative immune privilege of the cornea. The cornea is avascular, limiting access of lymphocytes and other immune responsive cells.[3] There are no associated lymph nodes; therefore, the opportunity for presentation of foreign antigen to antigen-presenting cells and T cells is also limited. The cornea expresses MHC antigens to a lesser extent than other tissues, contributing to immune privilege. However, this can be compromised by prolonged inflammation, extensive vascularization, and other factors, resulting in rejection.

The term graft rejection refers to the specific immunologic response of the host to the donor corneal tissue. Because it is a specific process, it should be distinguished from other causes of graft failure that are not immune mediated. A corneal graft that has suffered this immunologic response may or may not ultimately fail. Some physicians distinguish between graft reaction, which is reversible with medical therapy, and graft rejection, in which the immunologic end stage has been reached and the process is irreversible. Other physicians simply use graft rejection to refer to this immunologic process at any stage of its development, noting that some cases progress to graft failure because of rejection. This second terminology is used in this article because it is in line with terminology used in other types of organ transplantation.

Furthermore, at the time of presentation, determining with certainty whether an immune process is reversible is impossible.

Ritter et al discuss the need for further study of the genetic modification of corneal grafts prior to surgery to prevent rejection.[4]



United States

In 2016, 82,994 corneal transplantations were performed in the United States.[1] The incidence of graft rejection varies widely depending on the study design, type of transplantation, and risk factors for rejection. Reported incidences of penetrating keratoplasty graft rejection range from 2.3% to 68%.[5] At 5 years’ follow-up in the Cornea Donor Study, 23% of subjects had at least one rejection event, and 37% of the eyes with a rejection event had graft failure.[6]

Corneal graft rejection is the most common cause of graft failure in the late postoperative period.[2] The reported incidence of graft rejection is lower in endothelial keratoplasty. Descemet membrane endothelial keratoplasty (DMEK) has been reported to have a rejection rate as low as 0.7% at one year in one series, but other studies have reported higher rates.[7, 8]


Corneal graft rejection is the most common cause of graft failure in the late postoperative period.


No difference in corneal graft rejection between different races is known.


No sex predilection for corneal graft rejection is known.


Host age may influence the risk of corneal graft rejection. Some investigators have concluded that hosts older than 60 years have a lower risk of corneal graft rejection, although this has not been confirmed. The effect of donor age on corneal graft survival has been debated. The Cornea Donor Study did not find an association between donor age and corneal graft survival among corneal transplants at moderate failure risk.[9] Infants have higher rates of graft rejection than adults.


The sooner an episode of graft rejection is detected clinically and therapy is begun, the better the prognosis for graft survival. The rate of reversal of corneal endothelial graft rejection has been reported from 50-91%, depending on the clinical setting. In general, the prognosis is good if therapy is immediately instituted.

Depending on the degree of irreversible damage to the graft endothelium, even markedly edematous grafts may clear again. Once endothelial destruction has progressed to the point where the remaining endothelial function is inadequate to maintain deturgescence, the graft fails and becomes irreversibly edematous. Unfortunately, the endothelium has no or at best a very limited capacity for repair through mitosis.

Patient Education

No symptoms are related universally to graft rejection.

Astute patients may complain of a decrease in visual acuity, redness, pain, irritation, and photophobia. Patients may also be asymptomatic.

Any patient with a corneal graft should be instructed to seek ophthalmologic care urgently if these symptoms persist for more than a few hours.




Diagnosis of corneal graft rejection should be made only in grafts that have remained clear for at least 2 weeks following keratoplasty. By observing this guideline, graft rejection can be easily distinguished from other causes of graft failure that are more common in the early postoperative period (eg, primary donor failure). A sensitized host may rarely exhibit immunologic graft rejection before this 2-week period. Graft rejection has been observed to occur as late as 20 years after transplantation. Incidence of graft rejection is greatest in the first year following transplantation. A study of corneal graft surgery in the UK from 1999 to 2009 found endothelial failure as the most frequent indication for keratoplasty.[10]

No symptoms are related universally to corneal graft rejection, although astute patients may complain of the following:

  • Decrease in visual acuity

  • Redness

  • Pain

  • Irritation

  • Photophobia

Depending on the severity of the graft rejection, patients may be asymptomatic. Any patient with a corneal graft should be instructed to seek ophthalmologic care if these symptoms occur for more than a few hours.


Animal models of graft rejection reveal that the 3 corneal layers, epithelium, stroma, and endothelium, can be rejected separately. Although these separate rejection processes have been observed in humans, many patients present with combinations of epithelial, stromal, and endothelial rejection.

Epithelial rejection

Epithelial rejection presents in one of two manners.

The first type is characterized by an irregular, elevated epithelial rejection line that stains with fluorescein or rose bengal. The rejection line progresses rapidly across the cornea over several days to 2 weeks. A variant of this presentation may occur in which the epithelial rejection line takes the form of a ring, concentric with the limbus, which begins peripherally at the graft-host junction and progresses by shrinking centrally to a point. The rejection line represents a region of destruction of donor epithelium; the resulting epithelial defect is covered by host epithelium that grows inward from the remaining host cornea and limbus to cover the graft.

The second type of epithelial rejection is characterized by the presence of subepithelial infiltrates. These infiltrates consist of leukocytes and frequently have an appearance similar to the subepithelial infiltrates seen in adenoviral keratoconjunctivitis. These lesions may change location and shape over time, and they generally disappear on their own after several weeks.

Both types of epithelial rejection are steroid responsive, but, in many cases, the patient is either asymptomatic or has symptoms only of minimal irritation. As a result, the patient may not present to the ophthalmologist during these episodes. Although epithelial rejection generally is self-limited and tends not to cause visual disturbance on its own, it should be treated when found on examination as it may herald a more severe endothelial rejection.

Stromal rejection

Generally, stromal rejection in humans accompanies endothelial rejection and is difficult to demonstrate alone. It is characterized by peripheral full-thickness haze with limbal injection in a previously clear graft. An arc-shaped infiltrate may be noted peripherally at the graft-host junction that progresses centrally.

Endothelial rejection

Classic endothelial rejection presents with an endothelial rejection line (Khodadoust line) that usually begins at a vascularized portion of the peripheral graft-host junction and progresses, if untreated, across the endothelial surface over several days. The rejection line consists of mononuclear white cells that damage endothelial cells as the line sweeps across the endothelium.

Generally, a mild-to-moderate anterior chamber reaction is present. The damaged endothelium is unable to properly dehydrate the corneal graft; as a result, the donor cornea is clear ahead of the rejection line and is cloudy and edematous behind it.

A second variant of endothelial rejection is more diffuse in character, with scattered keratic precipitates and an anterior chamber reaction indicative of endothelial rejection and damage. In this type of endothelial rejection, stromal edema is typically not localized, but rather generalized throughout the graft, consistent with the generalized endothelial damage. The combination of keratic precipitates, an anterior chamber reaction, circumcorneal injection, and regions of corneal edema should be diagnosed as corneal graft rejection. In some cases, it may be difficult to distinguish graft edema from rejection and graft edema from endothelial insufficiency. Because rejection may be reversible, treating patients as if they have graft rejection is best.[11]

Allograft rejection after DMEK

Allograft rejection after DMEK may have a different clinical course with a slower onset than after penetrating keratoplasty.[12]


A great deal of energy has been expended in trying to determine clinical risk factors for corneal graft rejection. Because corneal graft rejection is the leading cause of graft failure in the late postoperative period, being able to identify and treat those patients at highest risk for graft rejection is important. Unfortunately, patients undergoing corneal transplantation represent a heterogeneous population, and proving that certain factors uniformly increase the risk of graft rejection is difficult. Differences in study designs exacerbate these difficulties.

Risk factors for corneal graft rejection can be divided into host and donor risk factors.

Potential host risk factors include the following:

  • Corneal vascularization
  • Active inflammation or infection (“hot eyes”) or history of inflammation
  • Anterior synechiae
  • Indication for transplantation [13]
  • History of previous graft failure, herpes simplex keratitis, chemical burn, or trauma
  • Preoperative diagnosis of pseudophakic or aphakic corneal edema [6]
  • Prior ocular surgery, including glaucoma procedures or pars plana vitrectomy
  • Larger or eccentric grafts
  • Bilateral penetrating keratoplasty
  • Younger age of host (lower risk associated with age >60 years, higher risk in infants)

Potential donor risk factors include the following:

  • Human leukocyte antigen A (HLA-A), human leukocyte antigen (HLA-B), human leukocyte antigen DR (HLA-DR), and ABO blood type incompatibility [14]
  • Presence of donor epithelium upon transplantation
  • Pretransplantation corneal tissue media and preservation

One factor that has been decidedly proven to increase the risk of rejection is host corneal vascularization. Multiple studies have confirmed an increased risk of corneal graft rejection with increasing host vascularization, ranging from rates of 0-10% of graft rejection in avascular host corneas to rates of up to 25-50% in severely vascularized host corneas. The precise cause for this increased risk is believed to be the relative loss of immune privilege that accompanies the usually avascular central cornea.

Some of the risk factors listed above remain in some dispute. In some cases, multiple studies have yielded contradictory results, whereas, in other cases, an insufficient number of clinical studies exist. In all cases, these risk factors are modified by the particular clinical situation. In particular, studies regarding the role of the major histocompatibility complex and HLAs have yielded contradictory data, although several studies indicate a trend toward a decreased incidence of graft rejection occurring in matched corneal grafts. Currently, corneal grafts are not routinely HLA typed and matched in the United States, unlike other organ transplants. Further evidence is needed to justify the added cost and complexity of performing HLA typing prior to corneal transplantation.

The effect of donor age on corneal graft survival has been debated. The Cornea Donor Study did not find an association between donor age and corneal graft survival among corneal transplants at moderate failure risk.[9] Graft rejection rates are higher in infants than in adults.



Diagnostic Considerations

Consider corneal edema due to endothelial decompensation.

Differential Diagnoses



Laboratory Studies

Graft failure is a clinical diagnosis; no laboratory studies are useful in diagnosis.

Imaging Studies

Graft failure is a clinical diagnosis; no imaging studies are useful in diagnosis.

Other Tests

Specular endothelial microscopy can be performed to determine endothelial cell density, pleomorphism, and polymegethism.

Pachymetry can be performed to measure corneal thickness.



Medical Care

Treatment of graft rejection depends on the type of rejection; however, in all cases, topical corticosteroids are the mainstay of treatment. Epithelial or stromal rejection without endothelial involvement usually does not progress to graft failure. As previously noted, epithelial rejection may be a self-limited process. Nonetheless, epithelial and stromal rejection should be aggressively treated, because they indicate host immunologic recognition of the graft and may precede a more severe endothelial rejection. Topical corticosteroids (eg, dexamethasone 0.1%, prednisolone acetate 1%) are prescribed 4-6 times/d until the signs of rejection resolve, followed by a slow tapering of the topical medication. These patients should be followed closely to be certain that the signs of rejection are improving and that endothelial rejection has not developed.

In cases of endothelial rejection, treatment must be more aggressive if the episode is to be reversed. Topical corticosteroids should be used every hour while awake and as frequently as possible at night for 2-3 days, followed by every 2 hours while awake. Steroid ointment may be used at bedtime. Therapy should be continued until signs of rejection resolve. Topical medications should be tapered slowly over several weeks to a few months depending upon the patient's response to treatment. Therapy should be continued for at least 4 weeks in the absence of response before judging that the graft has failed.

Other routes of administration of corticosteroids can be used in more severe endothelial rejections, in recurrent rejections, or if the patient is at high risk (eg, alkali burns, patients with vascularized corneas). Corticosteroids may be given by subconjunctival injection (eg, dexamethasone phosphate 2 mg, betamethasone 3 mg in 0.5 mL). A less painful alternative is a collagen shield soaked in corticosteroids and applied to the cornea combined with frequent corticosteroid eye drops. The collagen shield results in a higher local concentration of steroid than can be obtained by the use of corticosteroid drops alone. The shield acts as a depot reservoir for the drug that slowly releases its contents during the period between topical applications. Higher steroid concentrations have been noted in the cornea, aqueous humor, iris, and vitreous, compared with hourly drops alone.

Systemic corticosteroids can also be used in cases of severe endothelial rejection. Oral prednisone is generally started at dosages of 60-80 mg daily and continued for as long as 1-2 weeks before tapering. In line with findings in other fields of medicine, data suggest that pulsed intravenous (IV) steroids may be more effective than oral prednisone in reversing corneal graft rejection. Pulsed steroids (a single IV administration of 500 mg methylprednisolone) have been shown to improve the percentage of graft survival compared with oral steroids in patients who present early (within the first 8 days) in a rejection episode. A nonsignificant trend toward improved survival in all episodes of rejection in favor of pulsed steroids exists. In addition, pulsed steroids reduce the risk of subsequent rejection episodes, which may be a significant benefit in higher risk corneal grafts. Pulsed steroids also avoid prolonged administration of oral steroids.

In all cases of rejection, intraocular pressure should be monitored closely, especially when frequent corticosteroids are used. If necessary, elevated intraocular pressure should be controlled by topical medications to prevent glaucoma and to improve the chance of graft survival.

Surgical Care

No surgical care has proven beneficial during an episode of acute graft rejection.

Some transplant surgeons scrape the donor corneal epithelium to reduce the antigen load.

No solid evidence suggests that removing the donor epithelium is beneficial in reducing the risk of subsequent graft rejection.

If an acute graft rejection episode progresses to graft failure, repeat corneal transplantation may be indicated, including penetrating keratoplasty or endothelial keratoplasty.


No dietary restrictions have been identified.


No activity restrictions have been noted.


Depending on the degree of injury sustained by the graft, graft rejection episodes can progress to graft failure due to rejection.

Long-Term Monitoring

Patients should receive close follow-up care with an ophthalmologist for corneal graft rejection.

Frontiers in Corneal Graft Rejection Therapy

Systemic immunosuppression with cyclosporine in high-risk corneal transplantation has yielded varying results.[13] Cyclosporin A is a potent immunosuppressive agent that has revolutionized transplant therapy by reducing rejection in heart, kidney, liver, and other organ transplants. Cyclosporine is a fungal protein that has a high degree of specificity for T-cell lymphocytes and inhibits T-cell–mediated immune responses. In some studies, systemic cyclosporin A (blood levels 130-170 mcg/L) has been shown to greatly increase the rate of graft survival in high-risk corneal transplantation when used prophylactically for 12 months following transplantation. Cyclosporin A therapy carries known significant risks, including hypertension, renal toxicity, hepatotoxicity, and neurotoxicity; it should be used only after a thorough medical evaluation. Careful postoperative monitoring is essential and is generally best completed in conjunction with other physicians.

Topical administration of cyclosporin A also has been examined[15] and has yielded conflicting results for both prophylaxis and treatment of graft rejection episodes. Substitution of topical cyclosporine for topical corticosteroids may decrease intraocular pressure in cases of postkeratoplasty ocular hypertension or glaucoma but result in an increased risk of graft rejection.[16, 17] Cyclosporin A is not readily able to penetrate the corneal epithelium. Differences in drug vehicle and corneal penetration may account for the different outcomes seen in the use of topical cyclosporin A. Interestingly, collagen shields impregnated with cyclosporin A increase the corneal penetration of cyclosporine and can successfully reverse graft rejection in rabbits. Although methods that improve corneal penetration of cyclosporin A may improve its efficacy, they also may increase its potential systemic adverse effects. Note that blood levels of cyclosporin A have been recorded after topical administration of cyclosporin A in olive oil.

Other potential agents in the treatment of corneal graft rejection include antimetabolites (eg, azathioprine, 6-mercaptopurine) and immunosuppressives (eg, tacrolimus [FK-506], rapamycin). Relatively few studies have been performed using these agents in corneal transplantation, and their role in corneal transplantation therapy has yet to be determined. Each of these medications is associated with significant systemic adverse effects.

Tacrolimus has received more study than the other agents. One group in England reported significant success in preventing and reversing corneal and limbal allograft rejection in high-risk eyes.[18, 19] They found that no patient with therapeutic levels of tacrolimus suffered irreversible graft rejection. Several patients suffered from systemic adverse effects, including irreversible renal failure. In the United States, tacrolimus has only rarely been used in the setting of corneal transplantation. In another study of patients with high-risk corneal transplants treated with systemic tacrolimus, 65% had clear grafts at 2 years.[20] Topical tacrolimus has also been used successfully for the treatment of graft rejection.[21]

One significant issue is that the cost of tacrolimus for this indication is not always covered.



Medication Summary

Corticosteroids are the mainstay of treatment of acute graft rejection. They can be given topically, via subconjunctival injection, or systemically.


Class Summary

These agents provide anti-inflammatory activity to suppress the natural immune response that leads to acute graft rejection.

Prednisolone acetate 1% (Omnipred, Pred Forte)

Most commonly used topical corticosteroid. Decreases inflammation by suppressing migration of polymorphonuclear leukocytes and reversing increased capillary permeability.

Dexamethasone (Decadron)

For various allergic and inflammatory diseases. Decreases inflammation by suppressing migration of polymorphonuclear leukocytes and reducing capillary permeability. Also used for subconjunctival injections.

Prednisone (Deltasone, Rayos)

May decrease inflammation by reversing increased capillary permeability and suppressing PMN activity.

Methylprednisolone (Solu-Medrol, Medrol, Depo-Medrol)

Decreases inflammation by suppressing migration of polymorphonuclear leukocytes and reversing increased capillary permeability. Single pulsed dose of IV steroids prior to PO steroids may improve final outcome.


Class Summary

Agents in this category inhibit key factors involved in the immune response. They may be used when graft rejection is not controlled adequately by systemic corticosteroids. 

Tacrolimus (Astagraf XL, Envarsus XR, Prograf)

Tacrolimus suppresses humoral immunity (T-cell activity). It is a calcineurin inhibitor with 2-3 times the potency of cyclosporine. Tacrolimus can be used at lower doses than cyclosporine can, but it has severe adverse effects, including renal dysfunction, diabetes, and pancreatitis. Levels are adjusted according to renal function, hepatic function, and adverse effects.

Cyclosporine (Gengraf, Neoral, Sandimmune)

Cyclosporine is a cyclic polypeptide that suppresses some humoral immunity and, to a greater extent, cell-mediated immune reactions such as delayed hypersensitivity, allograft rejection, experimental allergic encephalomyelitis, and graft-versus-host disease for various organs.