| Study | Study design | Indication | No. of Eyes | Follow-up, month | Other treatment | Findings |
|---|---|---|---|---|---|---|
| Iseli et al, 2008 [114] | Prospective case series | Infectious keratitis unresponsive to antibiotics | 5 | 1-9 | Topical and systemic antibiotic therapy | In all cases, progression of corneal melting was halted. Emergency keratoplasty was not required in any of the cases. |
| Micelli Ferrari et al, 2009 [115] | Case report | Bacterial keratitis caused by Gram negative E. coli | 1 | 1 | Topical and systemic antimicrobial therapy | Corneal edema almost completely resolved, corneal ulceration healed after 1 month |
| Makdoumi et al, 2010 [116] | Prospective case series | Infectious keratitis associated with corneal melting | 7 | 1-6 | Topical antibiotics (all except 1) | Corneal melting arrested and complete epithelialisation was achieved in all cases. |
| Moren et al, 2010 [117] | Case report | Suspected acanthamoeba keratitis | 1 | 9 | Broad-spectrum antibiotics | Rapid decrease of pain and necrotic material. Corneal reepithelialisation started within a few days and completed within 1 month. Complete wound healing after 2 months. BCVA improved from 20/1000 to 20/30 after 9 months. |
| Khan et al, 2011 [118] | Interventional case series | Acanthamoeba keratitis unresponsive to treatment | 3 | 2 | Multidrug conventional therapy | Rapid reduction in symptoms and decreased ulcer size after the first treatment session. Progress of improvement slowed after 1 to 3 weeks but renewed after the second application. Ulcers closed within 3 to 7 weeks of first application. In 2 patients, penetrating keratoplasty was subsequently performed for residual dense corneal scars. |
| Anwar et al, 2011 [119] | Retrospective case reports | Infective keratitis unresponsive to antimicrobial therapy | 2 | - | Antimicrobial therapy | Rapid resolution of infective keratitis, leaving residual stromal scarring. 1 patient required penetrating keratoplasty for residual dense corneal scars. |
| Makdoumi et al, 2012 [120] | Prospective non-randomised study | Bacterial keratitis | 16 | - | Antibiotics only given for 2 out of 16 eyes | All eyes responded to photochemical treatment. Improved symptoms, reduced inflammation. Epithelial healing achieved. One patient required human amniotic membrane transplant. |
| Price et al, 2012 [125] | Prospective, dual-center, interventional case series | Infective keratitis (bacterial, fungal, protozoan, viral) | 40 | - | Standard antibiotic treatment, 7 patients had previous keratoplasty | Keratitis did not resolve in 6 cases and penetrating keratoplasty was needed. CXL should be avoided in eyes with prior herpes simplex. CXL appeared most effective when infection depth was limited. Success higher for bacterial than fungal infections. |
| Kymionis et al, 2012 [121] | Case report | Intractable post-laser keratitis due to atypical mycobacteria | 1 | 3 | Maximum antibiotic therapy | All infiltrates and stromal edema resolved after 1 week. UDVA improved from counting fingers at 3 meters to 20/35. |
| Li et al, 2013 [122] | Prospective case series | Fungal keratitis unresponsive to treatment | 8 | - | Topical antibiotics | No complications noted. Hypopyon disappeared in all cases between 3 to 11 days after CXL. Healing of corneal epithelium and ulcer was achieved between 3 and 8 days after CXL. |
| Arance-Gil et al, 2014 [131] | Case report | Acanthamoeba keratitis unresponsive to medical treatment | 1 | 9 | Medical treatment | After CXL, symptoms and corneal appearance improved significantly but the ulcer did not heal completely. Patient required amniotic membrane transplantation and penetrating keratoplasty. |
| Saglk et al, 2013 [123] | Case report | Suspected fungal keratitis unresponsive to treatment | 1 | 6 | Extensive medical treatment | Epithelial defect disappeared and stromal infiltrate stayed inactive from 1 week to 6 months after the second treatment. |
| Shetty et al, 2014 [127] | Prospective case series | Microbial keratitis (bacterial and fungal) | 15 | - | Antibiotics / antifungals | 6/9 patients with bacterial keratitis and 3/6 patients with fungal keratitis resolved after CXL treatment. Patients with deep stromal keratitis or endothelial plaque failed to resolve. |
| Tabibian et al, 2014 [124] | Case report | Atypical fungal keratitis (Aureobasidium pullulans) | 1 | - | None | Corneal epithelium closed completely within 3 days and infiltrate was completely eradicated. |
| Said et al, 2014 [128] | Prospective clinical trial | Infectious keratitis with corneal melting (bacterial, fungal, amoebic) | 40; 21 case, 19 control | - | Case: Antibiotics + CXL Control: Antibiotics only |
Average healing time was 39.76 +/- 18.22 (PACK-CXL) and 46.05 +/- 27.44 (control). CDVA after healing was 1.64 +/- 0.62 (PACK-CXL) and 1.67 +/- 0.48 (control). The PACK-CXL group had a bigger corneal ulceration width and length. |
| Vajpayee et al, 2015 [129] | Retrospective case-file analysis | Moderate mycotic keratitis | 41; 20 case, 21 control | - | Case: Antibiotics + CXL Control: Antibiotics only |
Average healing time and final BCVA were similar in both groups. The additional CXL treatment did not have any advantage over medical treatment. |
| Uddaraju et al, 2015 [130] | Randomised clinical trial | Nonresolving deep stromal fungal keratitis | 13; 6 case, 7 control | - | Case: Antibiotics + CXL Control: Antibiotics only |
The trial was stopped due to a marked difference in the rate of perforation between the 2 groups. The CXL group had a significantly higher rate of perforation. |
| Bamdad et al, 2015 [132] | Prospective randomised clinical study | Moderate bacterial corneal ulcers | 32; 16 case, 16 control | 0.5 | Case: Antibiotics + CXL Control: Antibiotics only |
Mean treatment duration was 17.2 +/- 4.1 days in the case group and 24.7 +/- 5.5 days in the control group. Epithelial defects were smaller in the case group at 7 and 14 days. |