Therapeutic Penetrating Keratoplasty: A Retrospective Review

Purpose: To report the indications, outcomes, and complications of therapeutic penetrating keratoplasty (Th PK) in patients with corneal perforation and/ or nonhealing corneal ulceration. Methods: A retrospective review was conducted of 51 eyes of 51 patients undergoing Th PK between January 1, 2006 and April 15, 2016. Data collected included patient demographics, visual acuity (VA), size of the corneal infiltrate and epithelial defect, degree of corneal thinning/perforation, microbiological results, surgical details, and postoperative complications. Results: The average age at presentation was 56.0 years (range 6-92 years), and most of the patients were females (n=31, 60.8%). Th PK was performed for corneal perforation in 28 eyes (54.9% of cases), nonhealing corneal ulcer in 16 eyes (31.4% of cases), and imminent risk of corneal perforation in 7 eyes (13.7% of cases). Infection was the most common reason for performing a Th PK and was present in 92.3% (47/51) of all cases. Of the infectious cases, the most common etiologies were bacterial (44.7%, 21/47) and fungal (31.9%, 15/47). The most common identifiable risk factor for undergoing a Th PK was a history of contact lens wear, which was seen in 32.7% of patients. Initial anatomic success was achieved in all patients after performing Th PK. Most patients (33/51; 64.7%) had clear grafts at their last follow-up examination. There was an improvement in VA in 70.2% (33/47, where data were available) of the patients at the final postoperative visit compared with the preoperative visit. Average best postoperative VA (1.14 +/- 0.88 logarithm of the minimum angle of resolution [LogMAR]; 20/276) was significantly better than the presenting (1.98 +/- 0.68 LogMAR; 20/1910) and preoperative (2.18 +/- 0.55 LogMAR; 20/3,027) visual acuities (P<0.0001). The most common complication after Th PK was cataract, which was present in 81.8% (27/33) of phakic eyes in which lens status could be assessed, followed by graft failure (47.1%; 24/51), and secondary glaucoma (45.1%; 23/51). Five eyes developed infection in the therapeutic graft, four eyes had persistent corneal epithelial defect at their last follow-up visit, and two eyes underwent evisceration. Conclusions: Therapeutic penetrating keratoplasty achieves anatomic success and it is a useful procedure for restoring a stable cornea in cases in which infection fails to heal or when the cornea perforates. Furthermore, Th PK achieves corneal clarity and improves vision in most patients.