Topical administration of tacrolimus and corticosteroids with concentration gradients is effective in preventing immune rejection in high-risk keratoplasty: a 5-year follow-up study

BACKGROUND: To evaluate the ecacy of the topical administration of immunosuppressants and corticosteroids with concentration gradients in the management of patients with high-risk keratoplasty. METHODS: One hundred and six patients treated with topical immunosuppressants (50 eyes in the FK506 group and 56 eyes in the CsA group) and corticosteroid eye drops with concentration gradients were enrolled in the study. The rates of rejection episodes, irreversible rejection, graft survival, and related inuential factors were evaluated. RESULTS: The mean follow-up period was 48.1±7.9 months (range, 36-60 months). The rates of rejection episodes (P=0.043) and irreversible rejection (P=0.062) were 14.0% and 6.00% in the FK506 group and 37.5% and 7.1% in the CsA group, respectively. Kaplan-Meier survival analysis demonstrated a signicantly higher graft survival rate in the FK506 group (81.6%±5.3%, 71.1%±6.3%) compared with that in the CsA group (71.1%±6.3%, 57.5%±7.5%) at 3 and 5 years after surgery (P=0.006). Multivariate logistic regression revealed that poor medication compliance with a preoperative risk score ≥ 3 (P=0.016) and endothelial immune rejection (P=0.033) were risk factors associated with graft survival. CONCLUSIONS: Topical administration of tacrolimus and corticosteroids with concentration gradients is effective in decreasing the incidence of immune rejection in high-risk keratoplasty. Careful instruction of patients on the reasonable use of topical tacrolimus is critical to avoid immune rejection induced by sudden discontinuation of medication.

CONCLUSIONS: Topical administration of tacrolimus and corticosteroids with concentration gradients is effective in decreasing the incidence of immune rejection in high-risk keratoplasty. Careful instruction of patients on the reasonable use of topical tacrolimus is critical to avoid immune rejection induced by sudden discontinuation of medication.

Background
Immune rejection after corneal transplantation remains the leading cause of graft failure. The 5-year survival rates of corneal grafts decreases dramatically in high-risk keratoplasty, ranging between 25% and 65%. [1][2][3][4][5] Although corticosteroids are currently the mainstay of treatment for routine postoperative management, they are insu cient in preventing graft rejection in high-risk patients. 6 For such cases, a variety of systemic immunosuppressants have been administered to prevent or reverse immune rejection.
Nonethless, these medications can be associated with severe side effects such as nephrotoxicity, hepatotoxicity, hypertension, and altered glucose metabolism. [7][8][9] Thus, topical administration of immunosuppressants like tacrolimus (also named FK506) and cyclosporine (CsA) is preferred. But limited data exist on the e cacy of topical immunosuppressants in preventing immune rejection in highrisk keratoplasty. [10][11][12][13] In this study, we aimed to evaluate the e cacy of topical tacrolimus and CsA with 5 years of follow-up, and concluded topical administration of tacrolimus and corticosteroids eye drops with concentration gradients is effective in preventing immune rejection in high-risk keratoplasty.
Methods PATIENTS: Topical 0.1% tacrolimus eye drops (Senju Pharmaceutical Ltd) is not currently approved for use in the treatment of immune rejection. Accordingly, the potential bene ts and complications of offlabel topical tacrolimus treatment were fully described to, and written informed consent was obtained from, all patients before participation in the study. This study was approved by the Institutional Review Board of Shandong Eye Hospital and adhered to the tenets of the Declaration of Helsinki. 106 eyes (106 patients) undergoing high-risk keratoplasty at the Shandong Eye Hospital were recruited on a consecutive basis from Jan 2013 to Jan 2015. Preoperative risk score for all the patients were recorded according to the method reported by Sloper CM et al. 14 The patients were divided into two groups in accordance with their own wish, one is the FK506 group treated with 0.1% tacrolimus and corticosteroids eye drops, and the other is CsA group treated with 1% CsA ( Oral prednisolone (1 mg/kg) was then started daily and tapered over a period of 2 to 3 months. Topical 1% Prednisolone Acetate eye drops was used 4 times per day for 1 month. Afterwards, 0.1% uorometholone was used 4 times daily for 6 months and tapered to 0.02% uorometholone three times daily for at least 1 year. Tobramycin and dexamethasone ophthalmic ointment was administered every night for 6 months and tapered to twice weekly. [15][16][17] In addition, 0.1% tacrolimus eye drops or 1% CsA eye drops was given 4 times per day for 1 month and tapered to three times for 6 months and twice for at least 1 year.
ANTIREJECTION THERAPY: When immune rejection occurred, intravenous methylprednisolone (2 mg/kg) was given daily for 5 to 7 days. Oral prednisolone (1 mg/kg) was then started daily and tapered over a period of 1 to 2 months. Tobramycin and dexamethasone eyedrops were administered every 2 hours for the rst 3 days and tapered to 4 times per day for the next 2 to 3 weeks. Afterward, 0.02% uorometholone eyedrops were applied 4 times daily. Tobramycin and dexamethasone ophthalmic ointment was used every night for 1 month and tapered to twice weekly. [15][16][17] Meanwhile, 0.1% tacrolimus eye drops or 1% CsA eye drops were given 4 times per day. MAIN OUTCOME MEASURES: Patient history, demographic information, preoperative risk factors, onset time of immune rejection, symptoms, and medication compliance were collected. Complete ocular examinations were performed, including best-corrected visual acuity (BCVA), intraocular pressure and slitlamp examination.
STATISTICAL ANALYSES: All data were described as mean values ± standard deviation. Statistical analyses were performed using SPSS 22.0 (SPSS, Chicago, Illinois, USA). A P value of ≤ 0.05 was considered statistically signi cant. The demographics and preoperative risk score were compared with the Wilcoxon signed ranks test between the two groups. The rejection episodes and irreversible rejection (loss of graft transparency) in each group were analyzed using the Mann-Whitney U test. Kaplan-Meier survival analysis and log rank tests were performed to evaluate the graft survival. The in uential factors, including age, gender, preoperative risk score (≥ 3 or < 3), surgical treatment (penetrating keratoplasty or keratolimbal allograft), topical immunosuppressants (0.1% tacrolimus eye drops or 1% CsA eye drops), poor medication compliance with 0.1% tacrolimus eye drops or 1% CsA eye drops, type of immune rejection (endothelial and non-endothelial immune rejection) were analyzed using multivariate adjusted logistic regression.

Results
The mean follow-up period was 48.1 ± 7.9 months (range, 36-60 months). Sixty eight patients were male, and forty eight patients was female. The mean age was 49.7 ± 12.2 years (range, 11-70 years). The demographics and preoperative risk score of the two groups were comparable and intergroup comparisons showed no statistical differences (Table 1).
REJECTION EPISODES: In FK506 group, immune rejection was observed in 7 eyes, with stromal rejection in 3 eyes and endothelial rejection in 4 eyes, the rate of rejection episodes was 14.0% (7/50). The causative factors was stopping taking FK506 eye drops in 7 eyes (with the average interval of 4.9 ± 0.2 days). Rejection occurred within 6 months after surgery in 4 eyes, at 6 months to 1 year in 1 eyes, at 1 year to 2 years in 1 eye, and at 2 years to 3 years in 1 eyes. 4 eyes restored a clear graft after 7.9 ± 1.40 days of antirejection therapy. However, the corneal grafts showed continuously edematous and opaque in 3 eyes. The rate of irreversible rejection was 6.00%.
In CsA group, immune rejection was observed in 21 eyes, with stromal rejection in 4 eyes and endothelial rejection in 17 eyes. The rate of rejection episodes was 37.5% (21/56), and the difference was statistically signi cant (P = 0.043) between the two groups. The causative factors included poor compliance with medications in 5 eyes, stopping taking corticosteroids eye drops in 6 eyes (with the average interval of 8.4 ± 2.3 days), and stopping taking CsA eye drops in 10 eyes (with the average interval of 6.3 ± 2.5 days). Rejection occurred within 6 months after surgery in 8 eyes, at 6 months to 1 year in 7 eyes, at 1 year to 2 years in 3 eye, and at 2 years to 3 years in 3 eyes. 17 eyes restored a clear graft after 8.8 ± 2.2 days of antirejection therapy. However, the corneal grafts showed continuously edematous and opaque in 4 eyes.
The rate of irreversible rejection was 7.1%, and there was no statistically signi cant difference (P = 0.062) between the two groups.
GRAFT SURVIVAL: The graft survival was 81.0%±7.4%, 72.0%±8.9% at 3 years after surgery, and 71.9% ±6.2%, 61.2%±6.9% at 5 years after surgery in FK506 group and CsA group, respectively (Fig. 1). Kaplan-Meier analysis and log rank tests showed that patients in FK506 group had a signi cantly higher graft survival rate both at 3, 5 years after surgery than patients in CsA group, and the difference was statistically signi cant (P = 0.000).   (14 eyes) sensation on drug instillation, which were more often occurred in CsA group (redness in 8 eyes, burning in 9 eyes, and stinging in 9 eyes). No cataracts or elevation of intraocular pressure were detected in the two groups during the follow-up.

Discussion
High-risk keratoplasty have been de ned as having at least two quadrants of stromal vascularization and/or a history of previous graft rejection. Other risk factors include chemical burn, corneal grafts diameter exceeding 9 mm, perforation or ocular in ammation at the time of surgery, and low recipient age. 13,14,18,19 Immune rejection in high-risk keratoplasty remain a therapeutic challenge to eye doctors.
They made lots of efforts on variety and usage of anti-rejection medications, 4,11,12,20 but still lacked an ideal treatment strategy. In our series, we applied 0.1% tacrolimus and corticosteroids eye drops with concentration gradients to patients with high-risk keratoplasty, and concluded it is effective in reducing immune rejection and prolonging graft survival.
Corticosteroid therapy remain the mainstay of preventing corneal graft rejection, for its dramatic inhibition of dendritic cell (DC) differentiation and maturation, and restoring a nonin amed microenvironment to support the transplanted graft. 5 But the regimen varies widely among respondents in the Cornea Society survey. Long-term use of corticosteroid eye drops is not recommend due to its underlying side events such as increased intraocular pressure or cataracts. Given that the peak time of immune rejection was 1 to 3 months after high-risk keratoplasty, 15 intensive topical and intravenous steroids were given within 1 mouth after surgery, and tapered over a period of 2 months. Afterwards, uorometholone was prescribe and tapered to a maintenance dose. Fluorometholone eye drops rapidly form inactive metabolites in the corneal tissue, and only a small portion passes through the cornea into the aqueous humor, 21 thus reducing the possibility of side effects associated with elevations in intraocular pressure or cataracts. Unlike the results of Zhai et al, rejection was observed after 30 months after surgery in 4 eyes in our study, making irreparable loss of graft endothelial cell. 22 Therefore, we advocate maintenance dose of corticosteroid eye drops during the follow-up period, and regular detection of intraocular pressure is worthwhile.
Tacrolimus and CsA are calcineurin-blocking drugs that inhibit clonal expansion of T lymphocytes through binding of intracellular proteins called immunophilins. 20 Topical CsA has been prescribed for years to treat different immune diseases of the eye. However, majority of prospective studies have failed to demonstrate any bene t with the use of topical CsA for high-risk keratoplasty. 4,12,23 Tacrolimus has been extensively used in preventing immune rejection for human organ transplantation. But few case series have reported bene cial effect of topical tacrolimus in human high-risk corneal transplantation. In this study, the rate of rejection episodes in FK506 group was much lower than that in CsA group (P = 0.043). Moreover, the graft survival was signi cantly higher both at 3, 5 years after surgery when comparing with the study conducted by Chow et al. 19 Systemic side effects on blood pressure, renal, and liver function were avoided consequently. A small sample size is one limitations of our study, which may in uence the treatment response. Further investigation on a larger sample seems to be necessary to evaluate the medication e cacy.
The e cacy of tacrolimus as an immunosuppressive agent is 10-100 times higher than that of CsA, 20 but sudden discontinuation of tacrolimus is more likely to induce immune rejection, with an average interval of 4.9 ± 0.2 days (shorter than that of CsA), which is a problem that can not be ignored. In this study, the causative factors for immune rejection included poor compliance or stopping taking medications without doctors' recommendations, thus careful medicine instructions to patients can be helpful. The restoration of corneal clarity after immune rejection is directly related to the interval from symptom onset to treatment and the degree of immune response. The longer the interval is and/or the severer the endothelial rejection is, the harder it will be to restore corneal transparency (P = 0.033). In addition, there was no statistically signi cant difference in the rate of irreversible rejection (P = 0.062) between the two groups. This is consistent with the reported study by Hashemian et al, in which topical tacrolimus can decrease the recurrence of rejection. However, it may not improve rejection reversal success.

Conclusion
In conclusion, topical administration of tacrolimus and corticosteroids with concentration gradients is effective in decreasing the incidence of immune rejection, and signi cantly prolonging graft survival in high-risk keratoplasty. It is critical to recommend reasonable use of topical tacrolimus to patients, thus avoiding inducing immune rejection by stop taking in sudden.  Kaplan-Meier curve of graft survival rate of FK506 group and CsA group at 3, 5 years after surgery.