**Background:** Infectious keratitis (IK) is a leading cause of corneal blindness worldwide. Standard treatment with broad-spectrum antimicrobials is increasingly compromised by antimicrobial resistance, and complications such as corneal melting and perforation often require surgical intervention. Corneal cross-linking (CXL), originally developed for keratoconus, increases corneal rigidity and resistance to enzymatic digestion. When used for IK, it is termed PACK-CXL (photo-activated chromophore for keratitis–corneal cross-linking). The microbicidal effect of photo-activated riboflavin is well known, but the mechanical contribution of increased stromal rigidity to infection localization has not been isolated. This study aimed to prove the mechanical role of CXL in guarding against progression of IK and organism penetration, even in the absence of the riboflavin microbicidal effect.
**Methods:** This prospective interventional case-control experimental study included ten male New Zealand albino rabbits (1.8–2.7 kg) from the animal house of Kasr Alainy Hospital, Cairo University. Rabbits were free from anterior segment conditions and had negative conjunctival swabs. The right eye of each rabbit underwent central 8 mm epithelial removal and CXL using riboflavin 0.1% solution (10 mg riboflavin-5-phosphate in 10 mL of 20% dextran-T-500) applied every 3 min for 30 min, with UVA (360 nm, 3 mW/cm²) for 30 min (fluence 5.8 J/cm²). After CXL, ofloxacin 0.3% was applied three times daily until complete re-epithelialization. The left eye served as control. After 6 weeks, anterior segment OCT (ASOCT) was performed on cross-linked eyes. The next day, bacterial keratitis was induced in all 20 eyes. Isolates of Staphylococcus aureus (MRSA ATCC 33592) and Pseudomonas aeruginosa (MDR ATCC BAA-2108) were obtained from the medical microbiology department. Suspensions were prepared by culturing a single colony in nutrient broth for 24 h at 37°C. After creating linear epithelial injuries with a 21-gauge needle, intrastromal injection of pure bacterial suspension was performed: 10 eyes (5 rabbits) received P. aeruginosa, and 10 eyes (5 rabbits) received S. aureus. Slit lamp examination was performed every 12 h post-infection until killing, scored using the Hobden grading system (0–4 for seven parameters: conjunctival injection, chemosis, iritis, fibrin, hypopyon, stromal infiltrate, stromal edema; maximum score 28). Rabbits were killed within 72 h due to progression in control eyes. Specimens were preserved in 10% buffered formalin, stained with hematoxylin and eosin, and examined for inflammation, epithelial changes, ulceration, organism penetration, and depth of stromal affection.
**Key Results:** Prior to infection, all epithelial defects healed within 2–5 days. ASOCT at week 6 showed an evident demarcation line at the anterior third to two-thirds of the cornea. Before infection, all rabbits had normal corneas (score 0). After infection, the control group developed the severest inflammation, necessitating killing at 72 h. Corneal melting occurred in 6 eyes in the control group versus none in the cross-linked group. Hobden clinical scores (mean ± SD) at 72 h: P. aeruginosa cross-linked (group 1A) 9 ± 1.41; S. aureus cross-linked (group 1B) 8.8 ± 0.83; control (group 2) 27.6 ± 0.54. Histopathological examination showed that in cross-linked eyes, inflammatory cell infiltration and lymphocytic aggregation were confined to the superficial stroma (upper 30–50%), with intact epithelium and minimal affection of iris and ciliary processes. In control eyes, there were marked corneal abrasions and erosions with dense inflammatory infiltrates digesting the basement membrane, invading more than 75% of stromal depth, with lymphocytic aggregation reaching the anterior chamber and more evident affection of iris and ciliary processes.
**Clinical Implications:** This study provides strong experimental evidence that prior CXL significantly limits the severity and depth of bacterial keratitis, even when the antimicrobial effect of riboflavin is presumably absent (since infection was induced 6 weeks after CXL). The mechanical role of CXL—increasing corneal rigidity and resistance to enzymatic digestion—appears to localize infection to the anterior stroma and prevent corneal melting. These findings support the inclusion of early PACK-CXL in the treatment algorithm for infectious keratitis, potentially reducing the need for therapeutic keratoplasty and addressing the growing problem of antimicrobial resistance. Limitations include the lack of microbiological evaluation of bacterial burden (colony-forming units) and the use of an animal model. Further clinical studies are warranted to confirm these benefits in human patients.