Refractory infectious keratitis is one of the leading causes of corneal blindness. Penetrating keratoplasty (PKP) is a primary approach to serious infectious keratitis patients who did not feasible to drug therapy. However, the comparative study on PKP outcomes of different pathogens-caused severe infectious keratitis is still lack at present.

This study was to evaluate the therapeutic effect of PKP on severe keratitis infected by fungus, bacteria, virus and acanthamoeba.

The clinical data of a total of 221 eyes of 210 patients who underwent PKP from severe infectious keratitis at Affiliated Hospital of Qingdao University from January, 2008 to January, 2014 were retrospectively analyzed and reviewed. The patients were grouped to fungal infected group (155 eyes), bacterial infected group (30 eyes), virus infected group (28 eyes) and acanthamoeba infected group (8 eyes). The follow-up duration was 6-24 months. The effective rate of treatment, eyes in different graded visual acuities, rejected rate and recurrence rate were compared among different groups.

The percentage of fungal keratitis was 70.1% (155/221), and that of bacterial keratitis, virus keratitis and acanthamoeba keratitis was 13.6% (30/221), 12.7% (28/221) and 3.6% (8/221), respectively. The effective rates of PKP were 88.5%, 86.7%, 92.8% and 75.0% in the fungal infected group, bacterial infected group, virus infected group and acanthamoeba infected group, respectively, showing insignificant difference among the groups (χ²=4.022, P=0.259). The postoperative best corrected visual acuity (BCVA) was better in the bacterial infected group than that in the fungal infected group or virus infected group (Z=-5.125, -7.504, both at P<0.001), and the postoperative BCVA in the virus infected group was superior to bacterial infected group (Z=-3.189, P =0.001). The grafts were clear in 170 post-PKP eyes, and corneal opacity and neovascularization were observed in 26 eyes, and a few new blood vessels were seen in another 25 eyes after PKP. The number of eyes in different degree transparences was not significantly different among the groups (P=0.325). The rejected rate of grafts was 39.4% in the fungal infected group, which was significantly higher than 20% in the bacterial infected group and 17.9% in the virus infected group (P=0.043, 0.029). The recurrence rate of lesions was 20% in the fugal infected group, 3.3% in the bacterial infected group and 3.6% in the virus infected group, showing significant differences between fugal infected group and bacterial infected group (P=0.032) as well as between fugal infected group and virus infected group (P=0.033). Rejected response occurred within 3 months after PKP in 42.6% fungal-infected eyes, which was earlier than bacterial-and virus-infected eyes.

PKP is an effective treating approach to severe infectious keratitis. The postoperative BCVA is better in viral keratitis than that of bacterial keratitis or fungal keratitis. There is a higher rejection rate of grafts in fungal keratitis than that in bacterial keratitis or viral keratitis after PKP.

[Key words] Ulcer, corneal; Infection diseases; Keratoplasty/penetrating; Treatment outcome