Article Text

Download PDFPDF

Amniotic membrane graft to conjunctival flap in treatment of non-viral resistant infectious keratitis: a randomised clinical study
  1. Bahaa-Eldin Hasan Abdulhalim1,
  2. Mostafa Mohamed Wagih1,
  3. Ahmed A M Gad1,
  4. Ghada Boghdadi2,
  5. Ragy R S Nagy3
  1. 1Department of Ophthalmology, Faculty of Medicine, Zagazig University, Zagazig, Egypt
  2. 2Department of Microbiology and Immunology, Faculty of Medicine, Zagazig University, Zagazig, Egypt
  3. 3Department of Ophthalmology, El-Minia Eye Hospital, El-Minia, Egypt
  1. Correspondence to Professor Bahaa Hasan Abdulhalim, Department of Ophthalmology, Faculty of Medicine, Zagazig University, Zagazig, Egypt; bahaah2001{at}


Purpose To evaluate and compare the results of bipedicle conjunctival flap (CF) and cryopreserved amniotic membrane graft (AMG) in the treatment of non-viral infectious keratitis resistant to medical treatment.

Methods This prospective randomised interventional study included 40 eyes of 40 patients with resistant non-viral infectious keratitis. Twenty eyes received CF and 20 eyes received AMG. In the CF group, there were 12 eyes with fungal keratitis, 7 eyes with bacterial keratitis and 1 eye with Acanthamoeba keratitis. In the AMG group there were 13 eyes with fungal keratitis, 5 eyes with bacterial keratitis and 2 eyes with Acanthamoeba keratitis. In the CF group, three ulcers had descemetocele and four ulcers were perforated. In the AMG group, four ulcers had descemetocele and two ulcers were perforated. In CF, 360° peritomy was done and a bipedicle CF from the upper conjunctiva was dissected from Tenon's capsule, mobilised to cover the cornea and sutured to episclera. In AMG, one or two layers of AM were trimmed to fit the ulcer and sutured to the cornea. The follow-up period was 6 months.

Results Successful results were observed in 18/20 eyes (90%) in each group. Postoperatively, no significant differences between the two groups were found regarding success rate (p=1.0), epithelialisation time (p=0.75) or visual acuity improvement (p=0.84).

Conclusions CF and AMG are effective in treatment-resistant infectious keratitis. They could restore ocular surface integrity and provide metabolic and mechanical support for corneal healing. For large corneal perforation, it may be better to use another procedure such as penetrating keratoplasty to restore ocular integrity.

  • Cornea
  • Infection
View Full Text

Statistics from


Microbial keratitis is a sight-threatening condition. It may permanently affect corneal transparency and in severe cases it may lead to corneal stromal melting and perforation with severe complications such as glaucoma and endophthalmitis. Pathogens invading the cornea secrete various proteolytic enzymes and toxin that may directly destruct the corneal tissue.1 Also, release of proteolytic enzymes from polymorphonuclear leukocytes as a response to inflammation may increase corneal destruction.2

Prompt and aggressive medical treatment is needed to stop destruction of corneal tissue. Microbial keratitis resistant to medical treatment is a challenging condition. The use of conjunctival flap (CF) to treat chronic corneal ulceration was first described by Gunderson in 1958.3 Since then, many investigators have used CF to treat infectious keratitis resistant to medical treatment or in the presence of perforation.2 ,4–10

The use of human amniotic membrane (AM) for the treatment of epithelial defects with corneal ulcers was proposed by Kim and Lee in 1997,11 then many studies documented the usefulness of using AM transplantation (AMT) in the treatment of infectious keratitis and corneal perforation.12–15

The aim of this study was to evaluate and compare the surgical outcomes of CF and AMT in the treatment of non-viral infectious keratitis resistant to medical treatment or accompanied by descemetocele or perforation.


This institutional prospective randomised parallel-group study included 40 eyes with culture positive non-viral infectious keratitis. Culture negative and mixed infection cases were excluded. The patients were collected from the outpatient clinic of Zagazig University Hospital between June 2011 and June 2012. The study was approved by the Ethical Committee of Faculty of Medicine, Zagazig University. All procedures were done when written informed consent was obtained after explanation of the possible consequences of the treatment methods.


Demographic data were collected from the patients using a form (age, sex and any systemic disorders, past history of topical or systemic treatment). Also, a detailed history was taken, including history of contact lens wearing, corneal scratch, corneal foreign body and eye trauma.

Corneal scraping

Corneal scraping was done for all cases on suspicion of infectious keratitis. Patients who had not received any prior treatment had corneal scraping for culture and sensitivity at presentation. For patients who received antibiotic treatment, corneal scraping was done after the antimicrobial agents had been discontinued for 48 h. Scraping kits included a minimum of one slide for Gram staining and one agar plate for aerobic incubation. Corneal scraping was done under surface anaesthesia (0.5% proparacaine HCl eye drops) with slit lamp magnification. The surface of the ulcer was gently but firmly scraped with Bard Parker number 15 blades or a Kimura platinum spatula. The base and the leading edges of the ulcer were scraped from the periphery to the centre. The scraped materials were sent for laboratory studies to do culture and sensitivity tests.

Preoperative examination

All patients underwent ophthalmic examination. Symptoms like pain, redness, photophobia, watering and discharge were recorded. Examination of the eye was done focusing on visual acuity testing, and slit lamp examination of the conjunctiva, cornea, anterior chamber, iris, pupil and lens. Hypopyon size, ulcer size, ulcer site, ulcer staining and corneal infiltrates were recorded. Associated ocular signs like blepharitis, dacryocystitis, dry eyes and corneal sensations were also noted.

Diagnosis of infectious keratitis

Microbiological study was done at the Microbiology and Immunology Department, Faculty of Medicine, Zagazig University. The diagnosis of infectious keratitis was based on the result of a positive culture obtained by corneal scraping at admission. The scraped material was smeared on glass slides for Gram stain and Giemsa stain. The scraped material was also inoculated directly onto sheep blood agar, chocolate agar, nutrient agar, Sabouraud's dextrose agar, potato dextrose agar, thioglycolate, and brain–heart infusion broth. Sabouraud's and potato dextrose agar plates were incubated at 25°C to enhance the growth of fungi and the remainders were incubated at 37°C. Blood agar plates were incubated under aerobic and anaerobic conditions; chocolate agar was incubated with 5% carbon dioxide.

Preoperative medical treatment

Freshly prepared fortified eye drops (cefuroxim 50 mg/mL and gentamicin 14 mg/mL) were applied as starting treatment. In the first 3 days, eye drops were applied round the clock followed by drops every 2 h during waking hours until results of laboratory investigation were available.

For fungal keratitis, medical treatment with antifungal agents was initiated when definitive diagnosis of fungal keratitis was made. Antifungal agents, either 5% natamycin (Alcon Laboratories, Fort Worth, Texas, USA) or 0.15% amphotericin B (Bristol-Myers Squibb, New York, USA) were used.

For bacterial keratitis, if there was no clinical improvement with the starting treatment, the treatment was shifted according to the results of culture and sensitivity tests when available.

For Acanthamoeba keratitis, polyhexamethylene biguanide 0.02% (GlaxoWellcome, England) eye drops with oral ketoconazole 400 mg/day were used.

Indications for surgical intervention

The indications for operative interference were deep ulcer (>50% stromal loss with poor re-epithelialisation), descemetocele or corneal perforation; and medical treatment failure (no improvement after 2 weeks from intensive medical therapy).

Eligible patients were randomly assigned (on a computer-based selection) to receive either CF (CF group) or AMT (AMT group).

Operative management

The operations were done under peribulbar block supplemented with intravenous sedation, except in cases of corneal perforation, descemetocele or patients who were uncooperative, when general anaesthesia was used.

CF operation

A bipedicle CF was done. Peritomy was done for 360°. The superior conjunctive was incised about 8 mm from the limbus, undermined carefully separating it from the Tenon's capsule. The conjunctival incision was extended to create a bipedicle flap. The corneal epithelium and the necrotic tissue were shaved with a scalpel, then the CF was mobilised and secured to the episclera with 10/0 nylon sutures covering the cornea, then applying topical antibiotics eye drops.

Amniotic membrane graft operation

Human AM was prepared and preserved as previously described.16 ,17 The AM was trimmed to fit the corneal ulcer and was placed with its epithelium (basement membrane) side up, secured with interrupted 10/0 nylon sutures, with the suture knots buried. Double-layered AM (both with the epithelium side up) was performed in patients with corneal perforation or descemetocele.

The operated eye was patched for 6 h, and then administration of the medical treatment agents was resumed.

Postoperative care

Medical treatment was continued after surgery. Postoperatively, patients were followed up daily for the first week, weekly for the first month, and then monthly for 6 months. Routine ophthalmic examination was done with special attention to the cornea for signs of healing and presence of any complications.

The main outcome measures were location, size and depth of the lesion, epithelialisation time and persistence of infection. Secondary outcome measures included visual acuity and other complications.

Statistical analysis

The data were processed using the Statistical Package for Social Science (SPSS) V.13 (Chicago, Illinois, USA). Continuous variables were expressed as means ± SD and compared using Student's t test. Categorical variables were expressed as percentages and were analysed using the χ2 test. p<0.05 was considered statistically significant.


Forty eyes with infectious keratitis from 40 patients were eligible for surgical interference. Twenty eyes were assigned to the CF group and 20 eyes were assigned to the AMT group. The demographic data and the preoperative and postoperative characteristics for the CF group are shown in table 1. Fungal keratitis was found in 12 cases, bacterial keratitis was found in 7 cases and Acanthamoeba keratitis was found in 1 case. Postoperatively, 18 cases showed good epithelialisation while two cases showed recession of the flap (one after 5 days and one after 6 days) and were referred for immediate therapeutic keratoplasty and were excluded from further statistical studies.

Table 1

Demographic data and preoperative characteristics for conjunctival flap group

The demographic data and the preoperative and postoperative characteristics for the CF group are shown in table 2. Fungal keratitis was found in 13 cases, bacterial keratitis was found in 5 and Acanthamoeba keratitis was found in 2 cases. Postoperatively, 18 cases showed good epithelialisation while 2 cases showed melting of the AM graft (one after 4 days and one after 8 days) and were referred for immediate therapeutic keratoplasty and were excluded from further statistical studies. In one eye, the AM graft dissolved after 7 days and another AM graft was done with successful outcome. The AM graft dislocated in one eye after 5 days and was resutured with successful outcome.

Table 2

Demographic data and preoperative characteristics for amniotic membrane transplantation group

Table 3 shows a comparison of the studied parameters between the CF group and the AMT group. There was no statistically significant difference between the two groups regarding age (p=0.27), sex (p=0.20), affected eye (p=0.75), type of pathogen (p=0.74, 0.49 and 0.55 for fungus, bacterial and Acanthamoeba keratitis, respectively), ulcer size (mean diameter, calculated from maximum diameter and perpendicular diameter to the maximum diameter) (p=0.36), ulcer site (p=0.43), presence of descemetocele (p=0.68) or presence of perforation (p=0.38). The mean epithelialisation time in the CF group was 13.16±7.35 days (range 5–26 days), and in the AMT group it was 13.94±7.30 days (range 5–27 days). There was no statistically significant difference in epithelialisation time between the two groups (p=0.75). Also, there was no statistically significant difference in improvement of visual acuity at 6 months postoperatively (p=0.84).

Table 3

Demographic data for conjunctival flap group and amniotic membrane transplantation (AMT) group


To our knowledge, this is the only report that compares AMT and CF in treating infectious keratitis. In this study, we found fungal keratitis is the most common form (25/40) of resistant non-viral infectious keratitis. This may be due to the area from which we collected the sample, which is mostly agricultural, delay in diagnosis, improper therapy and/or improper use of medications. Also this may be due to poor penetration of antifungal treatment into the cornea.18 Bacterial keratitis was found in 12/40 cases and Acanthamoeba keratitis in 3/40 cases.

Our results demonstrated successful results of CF and AMT in managing non-viral infectious keratitis in 18 of 20 eyes (90%) in each group. Failure to establish epithelialisation occurred in two cases with CF; the two ulcers were perforated and caused by virulent organisms (Pseudomonas aeruginosa and Fusarium species). This also occurred in two cases in the AMT group; one ulcer was perforated and caused by P aeruginosa and the other one had descemetocele and was caused by Aspergillus species. Other ulcers with perforation (two cases in the CF group and one case in the AMT group) or descemetocele (two cases in each group) healed completely. It seems that failure of re-epithelialisation may be attributed to virulence of organisms and/or accompanied by a large perforation.

The beneficial effect of AM in treating infectious keratitis may be attributed to the composition of the membrane and its mechanism of action. AM acts as a biological bandage providing a suitable substratum for epithelial cell migration, promotes epithelialisation, inhibits inflammation and may have an antimicrobial effect.12 ,14 ,19 ,20 Solomon et al13 reported successful results in all four cases of AMT in treating non-viral infectious keratitis (one case with perforation and three cases with descemetocele), but they mentioned the type of pathogen in only one case (P aeruginosa). Successful results were also reported in treating 12 cases of severe bacterial keratitis with immediate maximal topical antibiotics followed by AMT at 48 h plus topical steroid treatment at 72 h.21 Chen et al15 (using AMT for treating fungal keratitis) have found complete epithelialisation in 12 eyes (75%) with active fungal keratitis. In our study, we found complete epithelialisation in 12/13 (92.3%) cases of fungal keratitis treated with AMT.

In the CF procedure, we performed peritomy for 360° to facilitate mobilisation and suturing of the upper CF. In the postoperative period, the lower limbal conjunctiva attached to the limbus. This may provide a source of limbal stem cell. The CF promotes healing by providing a biological bandage and bringing in blood supply to the infected area that increases access to humoral and cellular immunity. Buxton and Fox2 reported successful results using CF in treating a series of patients with Pseudomonas corneal abscess. Alino et al8 also reported successful results using total and parcial CF in treating 14 cases with chronic ulcerative keratitis, with few complications. Cremona et al9 reported improvement of two cases with Acanthamoeba keratitis using deep lamellar keratectomy with a CF. Khodadoust and Quinter10 reported successful results using partial pedunculated CF in treating 11 cases with bacterial keratitis in a retrospective study.

Our results also showed that there was no significant difference in epithelialisation time or postoperative recovery between the CF and AMT groups. The maximum duration for re-epithelialisaion was 26 days in the CF group and 27 days in the AMT group, which seems reasonable for the severity of the ulcers. In treating fungal keratitis with AMT, Chen et al15 reported melting of the AM graft in 4/16 cases with active fungal keratitis but none in the 7 cases with inactive fungal keratitis and the the epithelialisation time was comparable to our results (range 6–26 days in 12 cases with active fungal keratitis). Kim et al12 using AMT have reported complete healing of ulcerative wounds of all 14 cases with infectious keratitis (9 bacterial, 3 Acanthamoeba and 2 fungal keratitis), but they did not define the epithelialisation time. The 100% success rate of their results may be due to using AMT when the activity of the pathogen was suppressed and improvement of the lesion was apparent.

We found no significant difference in visual acuity at 6 months between the groups. All eyes attained visual acuity of 6/60 or less, except one eye in the AMT group that had a final visual acuity of 6/36. Although CF precludes corneal transparency more than AM, it seems that the main cause of decreased vision is the resulting corneal opacity complicating the healed ulcer. All eyes would be prepared for corneal transplant to restore vision.

We did not observe notable side effects when using AMT or CF in the surgical procedure or in the postoperative period. In the 6-month follow-up period, there was no recurrence of infection. We observed reduction in ocular inflammation within the first few days of surgery and improvement of healing in most of our patients. We also noticed improvement in visual acuity in most patients.

Our results indicate that CF and AMT are effective as a surgical treatment for infectious keratitis after failure of medical treatment or if there is corneal perforation, as they could restore ocular surface integrity and provide metabolic and mechanical support for corneal healing. The use of AMT may have an advantage as it does not preclude corneal examination like CF, it does not interfere with limbal stem cells that may potentially affect the success of subsequent penetrating keratoplasty, and in addition, it preserves the conjunctiva which may be needed for another ocular procedure such as glaucoma filtering surgery. If the preserved AM is not available and the risk of transmission of diseases could not be abolished (as in the use of fresh membrane or improper preservation), it is preferable to use CF. For large corneal perforation, it may be better to use other procedures, such as penetrating keratoplasty for restoring ocular integrity.


We are grateful to all participants for cooperation and support of this study.


View Abstract


  • Contributors B-EHA, MMW and AAMG designed the study. B-EHA, MMW, AAMG and RRSN collected the data. All authors conducted the study. GB performed lab work. B-EHA performed statistical analysis. All authors contributed to analysis and interpretation of data. B-EHA drafted the paper. Critical revision of the article was done by B-EHA and GB.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethical Committee of Faculty of Medicine, Zagazig University.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost Jonas