Fungal keratitis is a serious cause of blindness worldwide, especially in developing countries. The risk factors are vegetative trauma, contact lens use and contact lens solution, ocular surface disease, topical steroid or antibiotic use and immunosuppressive systemic states, and so forth [
This retrospective study followed the tenets of the Declaration of Helsinki and was approved by the ethics committee of The Third People’s Hospital of Changzhou. Written informed consent was obtained from each participant. All patients with fungal keratitis involving midstroma, not responding to topical antifungal medications including natamycin and voriconazole, were enrolled in the study. The diagnosis of fungal keratitis was based on positive results of in vivo confocal microscopy (IVCM) (Figures
(a) Fungal ulcer located at the corneal center. The diameter of the ulcer was 5 mm, and it involved midstroma by slit lamp examination. (b) The depth of the ulcer was 320
(a) The fungal ulcer (size: 5 mm) located in the paracentral of the cornea. This case suffered hypopyon of about 1 mm. (b) AS-OCT shows the strong reflection of the infiltrate; the depth was about 392
The material thus obtained from scraping was used for direct microscopic examination using Gram’s stain and 10% KOH mount and also inoculated into Sabouraud’s dextrose agar, blood agar, and chocolate agar for culture and identification of species by standard microbiological procedures [
Topical (1%) and intrastromal voriconazole (50 ug/0.1 ml) were prepared in the hospital pharmacy. The voriconazole powder of 200 mg (VFEN, Pfizer, USA) was reconstituted with 20 ml of lactated Ringer’s solution (LR) to obtain 20 ml of clear concentrate containing 10 mg/ml of voriconazole. A 1 ml aliquot of this solution was further diluted with 19 ml of LR to a concentration of 0.5 mg/ml (50 ug/0.1 ml) for injection [
Topical natamycin sulfate (5%, Alcon) and voriconazole (1%, VFEN, Pfizer) were instilled every hour once fungal keratitis was conformed. The response to therapy was noted on slit lamp examination and defined as “no response to therapy” if there was no change in the size or depth of the ulcer or infiltrates and defined as “worsened” if there was an increase in size or depth of the ulcer or infiltrates, or perforation. If there was no response to the combined therapy for 2 weeks, intrastromal voriconazole (50 ug/0.1 ml) was performed around the lesion. In case of worsening or no response to the previous injection, intrastromal injection was repeated at an interval of 72 h. Corneal debridement was performed when gray infiltration and necrotic tissue were obvious even after the satellite lesion disappeared and the ulcer size diminished.
All intrastromal injections were performed under topical anesthesia (0.4% oxybuprocaine hydrochloride eye drops, Santen, Japan) in aseptic conditions using an operation microscope. The reconstituted voriconazole (50 ug/0.1 ml) was loaded in a 1 ml tuberculin syringe with a 30-gauge needle. With the bevel down, the needle was inserted obliquely in the uninvolved, clear area of the stroma to reach the infiltrate at the midstromal level. Voriconazole (0.05 ml) was injected in four divided doses around the infiltrate to form a drug deposit around the circumference of the lesion [
Corneal debridement was also performed under topical anesthesia (0.4% oxybuprocaine hydrochloride eye drops, Santen, Japan) in aseptic conditions using an operation microscope. The corneal lesion was removed layer by layer, using a sterile crescent knife (Mani, Japan) and 0.12-microtoothed forceps, until the residual cornea was smooth with no obvious infiltrate (Figure
After intrastromal injection and corneal debridement, the patients were continued on the previously mentioned topical antifungal regimen. Patients were examined daily, and the response to therapy was recorded on the slit lamp. The infection was considered resolved when there was complete healing of epithelial defect with resolution of corneal infiltrate and no hyphae found by IVCM. Topical antifungal therapy was continued for at least 2 weeks after the complete resolution of infection. Patients with impending perforation and progressive worsening of infiltrates were taken up for keratoplasty.
The data were analyzed using SPSS 19.0 statistical software. The continuous variables were presented as mean ± standard deviation, and the pair
The details of fourteen patients (8 males, 6 females) enrolled in the study are shown in Table
Presentation and final outcome of cases with recalcitrant fungal keratitis that received intrastromal voriconazole combined with debridement.
Number | Size of infiltrate (mm) | Size of ulcer (mm) | Location | Depth ( |
Initial BCVA | Organism isolated | Sensitivity of Vori | Inhibition zone (mm) | Intervention | Residual depth ( |
Duration for healing (d) | NV | Final BCVA | Astig (D) |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | 5.0 | 3.0 | Peripheral | 392 | 20/100 | Fusarium | R | 0 | INJ = 1 | 324 | 9 | Y | 20/25 | 1.0 |
2 | 4.5 | 3.0 | Paracentral | 280 | 20/160 | Fusarium | I | 12 | INJ = 1 | 408 | 13 | N | 20/25 | 1.75 |
3 | 5.5 | 4.0 | Central | 320 | 20/200 | Fusarium | S | 28 | INJ = 2 | 392 | 20 | N | 20/32 | 1.0 |
4 | 6.0 | 4.0 | Paracentral | 332 | 20/63 | Fusarium | S | 22 | INJ = 2 | 365 | 22 | N | 20/20 | 0 |
5 | 7.0 | 4.0 | Paracentral | 350 | LP | Fusarium | R | 10 | INJ = 2 | 500 | 24 | Y | LP | — |
6 | 3.5 | 2.0 | Peripheral | 188 | 20/40 | Alternaria | S | 21 | INJ = 1 | 476 | 10 | N | 20/25 | 0 |
7 | 8.0 | 5.0 | Paracentral | 392 | 20/2000 | Fusarium | S | 18 | INJ = 2 | 340 | 14 | Y | 20/50 | 1.5 |
8 | 5.0 | 3.0 | Peripheral | 270 | 20/2000 | Fusarium | S | 18 | INJ = 1 | 456 | 11 | Y | 20/32 | 2.0 |
9 | 4.0 | 2.0 | Peripheral | 236 | 20/50 | Alternaria | S | 21 | INJ = 1 | 500 | 7 | Y | 20/25 | 0.5 |
10 | 6.5 | 4.0 | Paracentral | 344 | 20/80 | Fusarium | S | 23 | INJ = 2 | 383 | 28 | N | 20/32 | 0 |
11 | 5.5 | 3.5 | Paracentral | 348 | 20/160 | Fusarium | R | 0 | INJ = 1 | 329 | 9 | N | 20/32 | 2.0 |
12 | 4.5 | 2.5 | Paracentral | 300 | 20/63 | Fusarium | I | 12 | INJ = 1 | 358 | 8 | N | 20/25 | 1.0 |
13 | 7.5 | 5.5 | Paracentral | 360 | 20/63 | Alternaria | S | 15 | INJ = 2 | 334 | 25 | Y | 20/25 | 1.0 |
14 | 5.0 | 3.0 | Paracentral | 304 | 20/100 | Fusarium | R | 0 | INJ = 3, LKP | — | 32 | — | 20/25 | 3.0 |
BCVA: best-corrected visual acuity; LP: light projection; Vori: voriconazole; R: resistance; I: intermediary; S: sensitive; INJ: number of intrastromal voriconazole injection; NV: neovascularization; Astig: astigmatism.
Of the 14 patients enrolled in the study, the satellite lesion in 12 patients and hypopyon in 2 patients disappeared after intrastromal voriconazole, without secondary infection or corneal perforation. After injection, the size of infiltration decreased significantly to (4.32 ± 1.10)mm (
The mean follow-up time was (37.14 ± 7.38) days, and healing time was (15.38 ± 7.38) days. The residue corneal depth after healing was (397.31 ± 65.55)
Fungal keratitis is a vision-threatening infectious disease. It is managed mainly by antifungal agents. Keratoplasty or corneal transplant is usually reserved for acute management of corneal perforation and for visual rehabilitation following corneal scarring [
Intrastromal voriconazole has the potential to achieve adequate drug concentration at the site of infection through a targeted drug delivery process [
In our research, 0.05% voriconazole was injected in four divided doses around the infiltration to ensure formation of a barrage of intrastromal voriconazole around the entire infiltration. The size of infiltration diminished significantly through the mean 1.46 injections of intrastromal voriconazole, but the ulcer did not shrink in size. Though the sensitivity of intrastromal voriconazole to different fungal genus remains to be determined, we can conclude that intrastromal voriconazole as an adjunctive therapy may be undertaken in selected patients who are unresponsive to other forms of antifungal therapy.
The same concentration of intrastromal voriconazole and the need for repeat injections to attain good results have been reported in many studies [
In order to reduce the need for repeat injections and to accelerate corneal ulcer healing, corneal debridement was performed when the satellite lesion disappeared and the size of infiltration diminished (Figures
The debridement procedure could cause corneal lamellar rupture and a decrease in corneal biomechanical function. Moreover, this may lead to iatrogenic keratectasia or corneal melting. The deeper and larger of debridement, the greater the loss of corneal function, and the higher the risk of complications. The key of success is to select suitable candidates. In our research, we estimated the depth of infiltrate and hyphae involved by AS-OCT and IVCM prior to debridement to determine whether debridement was suitable and the optimal depth of debridement. In addition, the organism-cultured outcomes were aided in determining whether debridement was feasible. We deemed that debridement is more effective in treating lesions with horizontal growth direction, such as
In the available literature regarding intrastromal voriconazole for fungal keratitis, the BCVA improved differently [
Intrastromal voriconazole combined with cornea debridement as adjunct to topical drugs is secure and effective for treating recalcitrant fungal keratitis. Suitable candidate screening and a thorough estimate before treatment are important for success.
Anterior segment optical coherence tomography
Laser scanning
Best-corrected visual acuity
Lamellar keratoplasty.
The datasets analyzed during the current study are available from the corresponding author on reasonable request (Email: czdgh1975@sina.com).
This study followed the tenets of the Declaration of Helsinki and was approved by the ethics committee of The Third People’s Hospital of Changzhou.
Written informed consents were obtained from the patients for publication of any information contained within the manuscript itself. The copies of the written consent are available for review by the editor of this journal.
The funding body had no role in the design, conduct of this study, data analysis, decision to publish, or preparation of the manuscript.
The authors declare that they have no competing interests.
Yajie Sun collected the data and drafted this manuscript; Zhuo Sun participated in the collection of the data and technique support; Yukai Chen performed examination; Guohua Deng performed the surgery and reviewed this manuscript finally. All authors read and approved the final manuscript. Yajie Sun and Zhuo Sun contributed equally to the work presented here and should therefore be regarded as equivalent authors.
This study was supported by the Major Science and Technology Program of Changzhou Health and Family Planning Commission (ZD201511) and Science and Technology Support Program of Changzhou (CE20175039).