Long-term comparison of full-bed deep lamellar keratoplasty with penetrating keratoplasty in treating corneal leucoma caused by herpes simplex keratitis

Abstract

Purpose: To compare long-term outcomes of full-bed deep lamellar keratoplasty (DLK) with penetrating keratoplasty (PK) for treating corneal leucoma caused by herpes simplex keratitis (HSK).

Design: Retrospective, comparative, interventional case series.

Methods: setting: Institutional. patients: Inclusion criteria were patients with corneal scarring induced exclusively by HSK who underwent primary graft of full-bed DLK or PK and completed a minimum of 12 months postoperative follow-up. There was no significant difference of corneal scarring and vascularization between the 2 groups before surgery. Choosing PK or full-bed DLK depended on the patient's own willingness, except those patients with a preoperative endothelial cell count of less than 700 cells/mm(2) or whose endothelial cell count was undetectable were encouraged to undergo only PK. Exclusion criteria were patients with a past history of corneal perforation, nonprimary graft, non-HSK-related corneal scars, and failure to complete a minimum of 12 months of postoperative follow-up. Fifty-eight eyes of 58 patients in the full-bed DLK group and 63 eyes of 63 patients in the PK group met the inclusion criteria. main outcome measures: Postoperative managements, recurrence of HSK, graft rejection, graft survival rate, visual acuity, and corneal endothelial density.

Results: The mean postoperative follow-up duration was 45.8 ± 30.9 months in the full-bed DLK group and 47.9 ± 27.2 months in the PK group (P = .70). As compared with the PK group, the full-bed DLK group experienced earlier suture removal (P = .01), needed fewer postoperative visits (P < .001), and had a higher proportion of eyes with full withdrawal of oral acyclovir (P < .001) and topical corticosteroid (P < .001). There were a total of 21 episodes of recurrent HSK in the PK group, more frequent than the 7 episodes in the full-bed DLK group, among which recurrent epithelial keratitis amounted to 13 episodes in the PK group, remarkably more frequent than the 1 episode in the full-bed DLK group. Twenty-six eyes (41.3%) encountered rejection episodes in the PK group, but no rejection episode was found in the full-bed DLK group (P < .001). In 14 eyes in the PK group, graft failure developed because of graft rejection, recurrence of HSK, or both, whereas only in 1 eye in the full-bed DLK group did graft failure develop because of recurrence of HSK (P = .001). The clear graft survival rate in the full-bed DLK group was significantly higher than that in the PK group (P = .01). Corneal endothelial cell density was stable from 1 month through 5 years in the full-bed DLK group, but 51.3% cell loss was found in the PK group at 5 years after surgery. At the last visit, 66.1% of eyes with full-bed DLK grafts and 50.9% of eyes with PK grafts achieved a best-correct visual acuity of 0.5 or better (P = .10).

Conclusions: Advantages of full-bed DLK over PK are no allograft rejection, longer graft survival, earlier drug withdrawal of topical steroid and oral acyclovir, less recurrence of HSK, and fewer follow-up visits. Full-bed DLK is preferable for treating HSK-induced corneal scarring with relatively healthy endothelium and with no history of perforation.