Ocular Microbiology and Immunology Group
Back to OMIG Main Page

< Previous | 2026 Agenda and Abstracts | Next >

2026 OMIG Abstract

Risk Factors for Immediate and Delayed Therapeutic Penetrating Keratoplasty in Microbial Keratitis

Kamini Narendra Reddy¹, Shruti Anant¹, Manisha Acharya² Nupur Gupta², Gautam Parmar³,
Rashmita Ravishankar⁴, Josephine Christy⁴, and Nakul Shekhawat¹
on behalf of the International Multicenter Partnership for Advancing Corneal Transplantation

¹Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore, Maryland; ²Shroff Charity Eye Hospital, New Delhi, India; ³SNC Chitrakoot, Madhya Pradesh, India; ⁴Aravind Eye Hospital, Pondicherry, India

Purpose: To determine the prevalence, baseline predictors, and clinical characteristics associated with therapeutic penetrating keratoplasty (TPK) in microbial keratitis, and to compare characteristics of eyes undergoing immediate versus delayed TPK.

Methods: We conducted a prospective multicenter observational cohort study of eyes with microbiologically confirmed bacterial, fungal, Acanthamoeba, microsporidium, or Pythium keratitis or clinically/microbiologically diagnosed viral keratitis enrolled across 4 tertiary centers in India. TPK was classified as immediate (<48 hours of presentation) or delayed (>48 hours after presentation). Baseline demographic, microbiologic, and clinical features independently associated with TPK were evaluated using site-adjusted logistic regression.

Results: Among 5,248 eyes with microbial keratitis enrolled across 4 tertiary referral centers, 412 eyes (7.9%) underwent TPK, including 131 (2.5%) immediate and 281 (5.4%) delayed procedures. Compared with bacterial keratitis, polymicrobial infections were independently associated with TPK (adjusted odds ratio [aOR] 1.47), whereas fungal keratitis was not (aOR 1.20). The strongest site-adjusted predictors of undergoing TPK were infiltrate diameter >6 mm (aOR 30.80), epithelial defect diameter >6 mm (aOR 26.25), presenting visual acuity >1.0 logMAR (aOR 13.48), diffuse infiltrate (aOR 7.98), posterior stromal involvement (aOR 6.37), frank perforation (aOR 6.00), impending perforation (aOR 4.11), limbal involvement (aOR 3.84), stromal thinning (aOR 3.68), and endothelial plaque (aOR 3.44). Compared with immediate TPK, delayed TPK was more strongly associated with hypopyon (aOR 3.98), superficial plaque (aOR 3.46), fungal keratitis (aOR 2.80), feathery margins (aOR 2.07), satellite lesions (aOR 1.95), and worsening infiltrate despite therapy as the recorded indication for surgery (23.5% immediate TPK vs 33.8% delayed TPK; aOR 1.86). Conversely, stromal thinning (aOR 0.17), impending perforation (0.22), frank perforation (0.23), posterior stromal involvement (0.32), limbal involvement (0.32), diffuse infiltrates (0.40) and endothelial involvement (0.45) were all less likely in delayed than immediate TPK. Delayed TPK was performed a median of 14 days (interquartile range [IQR], 8 to 26 days) after presentation.

Conclusions: Approximately 1 in 13 eyes with microbial keratitis underwent TPK in this multicenter cohort. Immediate and delayed TPK demonstrated distinct clinical profiles. Immediate TPK was predominantly associated with severe structural compromise at presentation, whereas delayed TPK was associated with disease progression despite antimicrobial therapy, particularly fungal keratitis. Recognizing these distinct pathways to TPK may improve risk stratification, prognosis, and surgical decision-making.