Examines outcomes of repeat optical penetrating keratoplasty, suggesting a need for improved immunomodulation strategies.
Purpose: To determine the indications, practice pattern of immunomodulation (IM), and outcomes of repeat optical penetrating keratoplasty (OPK). Methods: Of 4454 OPKs performed from 2016 to 2023, 259 (5.81%) were OPK following previous failed OPKs (OPK/OPK) and these were included in the study; those with follow-up <1 m were excluded. Indications for primary, last graft, risk factors, and IM use were recorded. Hazard’s ratio of factors for graft failure was computed. Graft survival (clear graft at last follow-up) and graft survival time were analyzed. Results: Prior OPK failed twice in 69 (26.64%), thrice in 41 (15.83%), and >4 times in 13 (5.01%). Indications for repeat graft included rejection (83.78%), HSV (14.67%), and infection (13.90%). Pre-op IM was used in 18.91%, while post-op was in 7%. Younger age, prior therapeutic PK, corneal scars post infections, anterior segment dysgenesis, >2 quadrants deep vascularization, and >2 failed grafts increased risk of graft failure. Visual acuity improved at 1 month and declined after 3 months. 39% of grafts survived until last follow-up (mean 31.5 m), and the median survival time was 36 m (CI: 30.2 to 45.8 m). Conclusion: OPK/OPK subset constitutes close to 6% of all OPKs performed, and 39% of these regrafts remain clear over a mean follow-up of 31 m. Half regrafts fail 36 m post-OPK. More than 80% of these grafts fail post rejection, and one-third post infection. Males undergo repeat OPKs more. Younger age, deep vessels, previous infections, and >2 grafts pose higher risk of failure to subsequent grafts. Current IM use is limited to 7–20% with pre-existing high risk and needs further studies to enhance graft survival in these complex situations.
No takes yet. Share an insight, caveat, or question.
Gokhale et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: