FP616 : Evaluation of Cross-Linked Donor Corneas in Therapeutic Keratoplasty in Cases of Fungal Keratitis

Dr. Anjana Karunakaran, K19063, Dr. Namrata Sharma, Dr. Manpreet Kaur, Dr. Jeewan S Titiya

INTRODUCTION

Mycotic keratitisaccounts for approximately 1–44% of all cases of microbial keratitis, depending upon the geographic location^1,2. It is worldwide in distribution, but is more common in the tropics and subtropical regions. Common fungi causing keratitisinclude species of Aspergillus, Fusarium, Candida, Curvularia, and Penicillium.^1,2,3

The most common risk factor for fungal keratitis is trauma with vegetable material or objects contaminated with soil.^1,2Of the organisms that cause keratitis, fungi remain one of challenging organisms to treat. Fungal keratitis usually requires a prolonged course of treatment with antifungal agent(s) because of the fungistatic activity and poor bioavailability of these agents.

Once diagnosis of fungal keratitis is confirmed, medical treatment should begin immediately. The spectrum of the antifungals used for the treatment of fungal keratitis includes polyenes, azoles,echinocandins and allylamines.

Keratoplasty is primarily indicated for medical treatment failure. However, under certain situations such as; advanced keratitis, severe corneal thinning, impending perforations, keratitis threatening to involve limbus the decision to perform keratoplasty must be taken early. The incidence of microbial keratitis in corneal grafts has been reported to be as high as 11.9 percent in developing countries.⁴ The most common risk factors predisposing to graft infection are suture related problems and persistent epithelial defect .⁵ The other causes of postoperative infectious keratitis are  intraoperative contamination, recurrent host disease or poorly preserved donor cornea. The majority of reinfection occurs within first year with about half of them occurring in the first six months of surgery. While the most common organism detected in corneal graft infection is Staphylococcus epidermidis in developing world,¹¹fungal infections too can occur in countries with hot and humid weather and common causative organisms include aspergillus and candida.⁶Fungal infections tend to recur more frequently than bacterial incidence varying from 0 to 14 percent.

Increased resistance to enzymatic digestion by collagenase, pepsin, and trypsin in corneas subjected tocollagen cross linking (CXL) has been reported previously⁷.Riboflavin acts as a photo sensitizer and when exposed to ultraviolet A light (UVA) creates oxygen free radicals which in turn creates new links between collagen fibrils. This increases the biomechanical strength of the cornea. Oxygen free radicals interfere with the integrity of the bacterial cell membrane and hence there is an antibacterial activity.⁸

The collagen cross linking provides increased stromal tensile strength and resistance to enzymatic degradation, thus reducing the chances of melt. It causes ultra structural changes producing reduced susceptibility to microorganism and reduction in inflammatory cells and immune cells.⁹The antimicrobial effects of CXL have been attributed to two possible mechanisms. First, the known phenomenon of keratocyte apoptosis following CXL may also kill the pathogens, thereby inhibiting the progression of infection. Secondly, corneal melting is attained by pathogens via a process of enzymatic digestion and CXL has been demonstrated to increase tissue resistance to enzymatic digestion. Taken together, this suggests that CXL attributes greater resistance to corneal tissue against enzymatic activity, thus exerting its antimicrobial effect.

In this study we hypothesize that cross linked donor corneas will reduce the incidence of melt and reinfection in cases of therapeutic keratoplasty post infectious keratitis.

MATERIALS AND METHODS

The study was conducted as a randomized case control study with randomization done in accordance to a table of random numbers generated by the computer.

We recruited 53 eyes of fungal keratitis (sample size 40) in accordance to the selection criteria. The cases were randomly allotted to the cross linked and the untreated group.26 cases had donor corneas cross-linked in accordance to Dresden protocol prior to keratoplasty and in 27 eyes untreated corneas were transplanted. Patients were recruited after obtaining ethics committee approval. Written informed consent was taken. After the patients are selected based on the selection criteria i.e. inclusion/exclusion criteria, following work up of the patients was done based on study plan.

Selection Criteria         

The patients will be selected based on the following selection criteria.

Inclusion criteria:

Diagnosed case of fungal keratitis (KOH positivity or culture positivity) with an indication for penetrating keratoplasty

1. Involving >/= two thirds of corneal thickness
2. Size >/=6mm
3. Refractory to medical management.
4. Age>18 years
5. Fungal corneal abscess 

Exclusion Criteria:

1. Endophthalmitis
2. Active viral keratitis
3. One eyed patients
4. Unwilling to give consent
5. Responding to medical treatment
6. Recurrence ( previous graft/ CXL)
7. Age<18 years

The patients underwent detailed evaluation including clinical history,anterior segment evaluation and posterior segment evaluation.The patients underwent B-scan ultrasonography for posterior segment, scraping and confocal microscopy for etiological diagnosis . Donor corneas used in this study were all of non optical grade. Corneas were graded based on slit lamp evaluation and specular count.

After the complete evaluation, the patients randomly allocated to group I underwentkeratoplasty using cross-linked donor corneas\. The patients were followed up on first post-operative, one week, one month and three months.

The post-operative visual acuity, graft clarity, intraocular pressure and confoscan were done. Graft clarity was graded as Grade 4 if the grafts were optically clear with excellent view of iris details, Grade 2-3 (borderline) if there was moderate to significant corneal haze with or without good view of iris details, and Grade 1-0 (failed) for opaque grafts with poor view of iris and anterior segment details. ¹⁰

The patients were allocated randomly into two groups. For the group I patients, ex vivo cross-linking of human donor corneas was performed using Moria’s artificial anterior chambers. The corneas were de-epithelialized and cross linked according to Dresden protocol. Pretreatment with riboflavin solution (0.1% riboflavin/20% dextran) for 30 minutes irradiated with ultraviolet A (UV-A) light (l = 370 nm, irradiance = 3 mW/cm²) for thirty minutes using Avedro’s KXL System. After cross linking these corneas were used for penetrating keratoplasty. Patients of group II underwent penetrating keratoplasty with untreated corneas.

STATISTICAL ANALYSIS

Data was recorded in Microsoft Excel Spreadsheet. Categorical variables were summarized by frequency(percentage) while quantitative variables were summarized by mean + standard deviation(after confirming appropriate normal distribution).Days of symptomatology being non-normally distributed, was summarized by median(minimum, maximum).Chi square/ Fisher’s exact test, as appropriate, was used to compare frequencies between the two groups. Student’s “t” test was used to compare mean values between the two groups. Stata 14.0 statistical software was used for data analysis. All the statistical tests were two tailed. In this study, p value <0.05 was considered as statistical significant.

RESULTS

A total of 53 eyes were recruited in accordance to the inclusion criteria. There were 26 cases and 27 controls. The demographic profile of the patients was comparable andis summarized in table I.

Risk factors for development of fungal keratitis included trauma with vegetative matter (group I: 14, group II: 10), trauma with non-vegetative matter (group I: 4, group II: 4), previous intraocular surgery (group I: 3, group II: 3) and inadvertent topical steroid usage (group I: 1, group II: 0). No risk factor could be identified in 14 cases (group I:4, group II: 10).

The causative micro-organisms were Aspergillus species (group I: 12, group II: 14), Fusarium species (group I: 8, group II: 6), Candida species (group I: 2, group II: 3), Penicillium(group I: 2, group II: 1) and Alternaria (group I: 2, group II: 3).

The size of the ulcer was comparable in both the groups (Group I: 6.99 ± 1.36 mm, Group II: 7.01 ± 1.09 mm; p=0.965). Perforation was noted in 13 cases in group I and 15 cases in group II (p = 0.786).

The median interval between onset of fungal keratitis and keratoplasty was 20 days in both groups (range: 6 – 180 days in group I; 7-180 days in group II) (p = 0.99). The host bed trephination size was 8.48 ± 0.60 mm in group I and 8.15 ± 0.70 mm in group II (p=0.07). The donor button size was 9.49 ± 0.63 mm in group I and 9.15 ± 0.70 mm in group II (p=0.067). Majority of eyes were phakic (group I: 17, group II: 19), followed by aphakic (group I: 7, group II: 5) and pseudophakic (group I: 2, group II: 3).

Graft infection was observed in no case in group I during follow up (Table 2). In contrast, 6 cases had graft infection in group II and the difference between the two groups was statistically significant (p=0.011). Of these, three cases had recurrence of primary infections (Aspergillusfumigatus in two cases and Aspergillusniger in one case) and three cases had a new bacterial infection (Staphylococcus epidermidis in all three cases). The risk factors for graft infection were loose sutures with infiltrates in three cases, prior perforated corneal ulcer in 3 cases, secondary glaucoma in two cases, aphakia in 2 cases and diabetes mellitus in one case. The average time to reinfection was 14.67 ± 2.73 days. All cases received frequent instillation of fortified topical antibiotics/ antifungals and systemic antibiotics/antifungals based on the causative organism and sensitivity patterns. The infection resolved on conservative medical therapy alone in four cases, with an opaque graft and visual acuity ranging from 2/60-1/60. A repeat keratoplasty was performed in two cases with a recurrence of fungal infection; one eye eventually developed phthisis bulbi.

Graft clarity was significantly better in group I wherein 88.5% cases in group I had clear grafts as compared to 33.3% cases in group II (p<0.001) at the end of three months (Table 3). The graft host junction was well-apposed in all cases.

Postoperative CDVA was significantly better in group I at 1 and 3 months of follow up. The mean CDVA at the end of three months was 1.15 ± 0.35 logMAR units in group I and 1.69 ± 0.67 logMAR units in group II (p<0.001) [Table 4]. A total of 61.5% (16/26) cases had CDVA of 20/200 or better in group I at the end of three months, as compared to 29.6% (8/27) cases in group II (p= 0.0281).

The mean endothelial cell count was 1832 ± 37.13 cells/mm² in group I and 1842.963 ± 45.63 cells/mm² in group II (p=0.34). The stromal keratocyte density as assessed on confocal microscopy was significantly less in group I (group I: 1759.444 ± 32.56 cells/mm³, group II: 1841.333 ± 93.20 cells/mm³; p= 0.002).

All cases in group I had a total epithelial defect on postoperative day 1, as the donor cornea was de-epithelialized before cross-linking. The epithelial defect resolved by the end of first week in all eyes. In group II, 4 eyes had epithelial defect on first postoperative day, which resolved by the end of first week. A persistent epithelial defect was not observed in any eye. No eye in group I developed corneal melt. In group II, varying degrees of corneal melt were observed only in the six cases that developed infection.

Secondary glaucoma developed in 11.54% eyes (3/26) in group I and 25.92% eyes (7/27) in group II (p=0.18). The intraocular pressure was controlled by systemic and topical anti-glaucoma medications in all cases. Surgical intervention was not required in any case.

Loose sutures were observed in ten cases (group I: 4, group II: 6; p= 0.73). Of these, three cases in group II developed suture infiltrates and graft infection and were managed with fortified topical antibiotics. A suture replacement was done in the remaining cases.

Table 1: Demographic Profile of Cases of Fungal Keratitis Undergoing Therapeutic Penetrating Keratoplasty with Collagen-Cross Linked Donor Corneas or Non-Collagen Cross Linked Donor Corneas.

Demographic Parameter	Group I (TPKwith CXLtreated corneas) N=26	Group II (TPK with non-CXLtreated corneas) N=27	P value
Age (years)	50.46 ± 14.49	45.48 ± 16.16	0.244
Gender (Male: Female)	16:10	21:6	0.198
Laterality (Right: Left)	15:11	17:10	0.695

Table 2: Type of Infection, Causative factors, Management & Outcomes after Graft Infection following TPK

Case	Graft Infection (New/ Recurrent)	Causative Organism	Interval between TPK and Infection (days)	Risk factors	Management	Visual Outcome	Anatomical outcome
1	New	Staphylococcus epidermidis	13	DM, Loose Suture, Perforation	Topical fortified antibiotics* Systemic antibiotics#	1/60	Resolution of infection Opaque graft
2	New	Staphylococcus epidermidis	11	Loose Suture, Glaucoma	Topical fortified antibiotics* Systemic antibiotics#	1/60	Resolution of infection Opaque graft
3	New	Staphylococcus epidermidis	14	Loose Suture	Topical fortified antibiotics* Systemic antibiotics#	1/60	Resolution of infection Opaque graft
4	Recurrence	Aspergillusfumigatus	16	DM, Aphakia	Re-graft Topical & systemic antifungals$	Perception of light with inaccurate projection of rays	Phthsisbulbi
5	Recurrence	Aspergillusfumigatus	15	DM, Aphakia	Re-graft Topical & systemic antifungals$	1/60	Resolution of infection Opaque graft
6	Recurrence	Aspergillusniger	19	Perforation, Glaucoma	Topical & systemic antifungals$	2/60	Resolution of infection Opaque graft

 Infection following TPK

TPK: Therapeutic Keratoplasty; DM: Diabetes Mellitus

*Topical fortified Cefazolin 5% + Tobacin 1.3% every two hours

#Oral Ciprofloxacin 500 mg twice a day

$Topical Natamycin 5% every two hours and oral Ketoconazole 200 mg twice a day

Table 3: Postoperative Graft Clarity after TPK in Fungal Keratitis with CXL-treated and Non CXL-treated Donor Corneas

Postoperative Period	Group I (n=26)		Group II (n=27)		P value
	Clear grafts*	Failed grafts#	Clear grafts*	Failed grafts#
Day 1	9 (34.6%)	17 (65.4%)	7 (25.9%)	20 (74.1%)	0.491
Day 7	12 (46.2%)	14 (53.8%)	12 (44.4%)	15 (55.6%)	0.901
1 month	21 (80.8%)	5 (19.2%)	6 (22.2%)	21 (77.8%)	<0.001
3 months	23 (88.5%)	3 (11.5%)	9 (33.3%)	18 (66.7%)	<0.001

*Clear grafts were defined as graft clarity of 3 or 4

#Failed grafts were defined as graft clarity of 0, 1 or 2

DISCUSSION

Fungal keratitis is a complex entity with many considerations when it comes to diagnosis and treatment. It is particularly a public health concern causing visual loss in developing countries where there is a limited access to care and economic barriers. As with all corneal infections, proper identification of microbe and targeted therapy can reduce complications, Therapeutic keratoplasty is mainly indicated in cases of infectious keratitis refractive to specific antimicrobial therapy.¹¹The diseased tissue is removed and replaced with donor cornea, thus reducing the infectious load. Previous studies have shown PK as an effective option for treatment of refractory or severe cases of fungal keratitis.^12,13 The most common complications of therapeutic keratoplasty are reinfections, graft melting, secondary glaucoma, graft rejection, complicated cataract, endophthalmitis, persisting endothelial defect and anterior synechia.¹⁴Reinfection rates in therapeutic penetrating keratoplasty varies from 12-20%.¹⁵ There are various methods to reduce graft infection such as appropriate and complete removal of infected tissue, early therapeutic penetrating keratoplasty, proper suturing technique, post operative antimicrobial therapy and donor corneal cross linking⁴.Donor collagen cross linking reduces the susceptibility of the tissue to the organism due to ultra-structural changes in cornea. Chemical alteration of nucleic acids in residual microbes makes replication impossible. Increased stromal tensile strength and rigidity of corneal collagen prevents melting. It increases resistance to enzymatic degradation and there is reduced tendency to generate vascularisation. In our study, we did not observe any case of graft infection, either a new infection or a recurrence of primary fungal infection in the group that underwent TPK with CXL treated donor corneas. The difference in the incidence of graft infection was statistically significant between the two groups that underwent TPK with CXL-treated or untreated donor corneas respectively. The donor tissue quality was inferior to those used for optical penetrating keratoplasty, owing to a shortage of suitable donor tissues in developing countries, hence, the visual outcomes and graft clarity is suboptimal as compared to the outcomes of penetrating keratoplasty. Similar donor tissues were used in both groups, and the graft clarity and visual acuity were significantly better in the CXL group. The endothelial cell count of both groups was similar at the end of three months. The endothelial safety of CXL-treated human donor corneas has been established in experimental studies, wherein a similar endothelial cell count was observed in CXL-treated and untreated donor corneas.¹⁶ Confocal microscopy of the CXL treated corneal grafts revealed a compact stromawith a reduced keratocyte density as compared to non-CXL treated corneal grafts. The stromal keratocyte apoptosis in CXL treated corneal grafts reduces the immunological reaction and thus may have a protective effect against graft infection and melt. No case with CXL treated corneal grafts developed corneal melt. In non-CXL treated corneal grafts, one case with recurrence of infection progressed to total corneal melt and eventually developed phthisis bulbi. Remaining five cases with graft infection developed varying degrees of corneal melt, which necessitated repeat keratoplasty in three cases. In addition to increased resistance to keratolysis, CXL-treated donor corneas may be associated with additional advantages such as reduced induction of intraoperative astigmatism and aberrations, as well as a superior graft adhesion.^17,18,19 We observed well apposed grafts with no discontinuity in the graft host-junction in any case in both groups. No adverse effects such as persistent epithelial defect, endophthalmitis or phthisis bulbi were observed as a result of CXL.

CONCLUSION

Cross linking of donor corneas showed significant reduction in reinfection rate post therapeutic penetrating keratoplasty. Patients with cross linked donor corneas (group I) achieved better best corrected visual acuity and graft clarity at 3 months as compared to control group (group II).Further long term studies have to be conducted with a larger sample size in all cases of infectious keratitis to assess the effects of donor cross linking.

REFERENCES

1. Gower EW, Keay LJ, Oechsler RA, Iovieno A, Alfonso EC, Jones DB, et al. Trends in fungal keratitis in the United States, 2001 to 2007. Ophthalmology 2010;117:2263 7.
2. Garg P. Fungal, mycobacterial, and nocardia infections and the eye: An update. Eye (Lond) 2012;26:245 51.
3. Revankar SG, Sutton DA. Melanized fungi in human disease.ClinMicrobiol Rev 2010;23:884 928.
4. Chang, Steven D., et al. “Factors influencing graft clarity.” Cornea 15.6 (1996): 577-581.
5. Vajpayee, Rasik B., et al. “Infectious keratitis following keratoplasty.” Survey of ophthalmology 52.1 (2007): 1-12.
6. Harris, David J., et al. “Late bacterial and fungal keratitis after corneal transplantation: spectrum of pathogens, graft survival, and visual prognosis.” Ophthalmology 95.10 (1988): 1450-1457.
7. Iseli HP, Thiel MA, Hafezi F, Kampmeier J, Seiler T. Ultraviolet A/riboflavin corneal cross-linking for infectious keratitis associated with corneal melts. Cornea. 2008;27(5):590-4.
8. Makdoumi K, Mortensen J, Crafoord S. Infectious keratitis treated with corneal crosslinking. Cornea. 2010;29(12):1353-8.
9. Micelli Ferrari T, Leozappa M, Lorusso M, Epifani E, Micelli Ferrari L. Escherichia coli keratitis treated with ultraviolet A/riboflavin corneal cross-linking: a case report. European journal of ophthalmology. 2009;19(2):295-7.
10. McDonnell PJ, Enger C, Stark WJ, Stulting RD. Corneal thickness changes after high-risk penetrating keratoplasty. Collaborative Corneal Transplantation Study Group. Arch Ophthalmol 1993;111:1374-81.
11. Sharma, N., Chacko, J., Velpandian, T., Titiyal, J. S., Sinha, R., Satpathy, G., … & Vajpayee, R. B. (2013). Comparative evaluation of topical versus intrastromalvoriconazole as an adjunct to natamycin in recalcitrant fungal keratitis. Ophthalmology, 120(4), 677-681.
12. Xie L, Dong X, Shi W. Treatment of fungal keratitis by penetrating keratoplasty. Br J Ophthalmol. 2001;85:1070–1074.
13. Sharma, N., Agarwal, P., Sinha, R., Titiyal, J. S., Velpandian, T., & Vajpayee, R. B. (2011). Evaluation of intrastromalvoriconazole injection in recalcitrant deep fungal keratitis: case series. British Journal of Ophthalmology, 95(12), 1735-1737.
14. Sharma, N., Jain, M., Sehra, S. V., Maharana, P., Agarwal, T., Satpathy, G., & Vajpayee, R. B. (2014). Outcomes of therapeutic penetrating keratoplasty from a tertiary eye care centre in northern India. Cornea, 33(2), 114-118.
15. Boujemaa, C., et al. “Kératoplastietransfixante à chauddans les ulcèrescornéensinfectieuxperforés.” Journal françaisd’ophtalmologie 28.3 (2005): 267-272.
16. Mooren P, Gobin L, Bostan N, et al. Evaluation of UVA Cytotoxicity for Human Endothelium in an Ex Vivo Corneal Cross-linking Experimental Setting. J Refract SurgThorofare NJ 1995. 2016;32(1):41-46. doi:10.3928/1081597X-20151207-05.
17. Flavahan PW, McDaid K, Watters T, Finch M, Mantry S. Collagen cross-linking treatment increases adhesion in mock corneal grafts. Contact Lens Anterior Eye J Br Contact Lens Assoc. 2016;39(6):416-419. doi:10.1016/j.clae.2016.07.003.
18. Arafat SN, Robert M-C, Shukla AN, Dohlman CH, Chodosh J, Ciolino JB. UV cross-linking of donor corneas confers resistance to keratolysis. Cornea. 2014;33(9):955-959. doi:10.1097/ICO.0000000000000185.