top of page

Minimally invasive Keratoprosthesis (mi-KPro) for vision restoration in severe ocular chemically injuries.


Eleftherios Paschalis Ilios


Jyoti Sharma1, Thomas Dohlman1, Chengxin Zhou1, Peter York2, Fengyang Lei1, Jie Liu1, Robert J. Woods2, Claes H. Dohlman1, James Chodosh3, Roberto Pineda1, Eleftherios I. Paschalis1

1. Harvard Medical School - Mass. Eye and Ear

2. Harvard Engineering

3. University of New Mexico


Purpose: Severe ocular injuries often lead to corneal blindness. Artificial corneas have changed the prognosis and have become the mainstay in vision restoration. Still, long-term complications, such as secondary glaucoma, endophthalmitis, and retroprosthesis membrane (RPM) frequently occur. Here we present a new minimally invasive KPro (mi-KPro) that reduces such complications. It consists of elongated optical stem, uses duplicate Descemet’s membrane entry ports, and has an ultra-thin, flexible titanium backplate within the corneal stroma. Here we test the new design in models of corneal burns.

Methods: Severe corneal alkali or acid burns were achieved using 2N NaOH (n=12) or 2N HCL (n=12), respectively. One month later, injured corneas were restored either by using the mi-KPro or with standard penetrating keratoplasty (PKP) (control). Implantation of the mi-KPro device was achieved using deep anterior lamellar keratoplasty (DALK) for backplate placement and a 2.5mm diameter central Descemet membrane trephination for insertion of the optical stem into the anterior chamber. PKPs were performed using a 8mm trephine. A donor corneal graft was used in all surgeries. Eyes received temporary tarsorrhaphy for 7 days and followed monthly for 1 year using biomicroscopy, anterior and posterior segment optical coherent tomography (OCT), and intracameral pressure measurements. No eye drop prophylaxis was administered beyond the first 3 weeks of the surgery. Ocular tissues were collected for further histological analyses using hematoxylin and eosin and p-Phenylenediamine staining.

Results: The anatomic retention of the mi-KPro was 100% in 1 year after acid burn and 100% in 6 months after alkali burn in rabbits. In contrast, all standard PKP eyes (n=12) exhibited rapid graft failure within 4 months. The new mi-KPro design did not cause intraocular pressure elevation!, glaucoma, RPM, or infection at 1 year. In contrast, all PKP eyes exhibited refractory IOP elevation (p=0.004, n=12), thinning of the retina (n=6, p=0.004), and optic nerve degeneration (n=6, p=0.01) within 4 month of the surgery.

Conclusion: Our data suggest that mi-KPro achieves excellent anatomic retention in chemically injured rabbit eyes for a year, does not lead to RPM formation, and most importantly, does not cause IOP elevation or glaucoma. This design is ready for clinical trials.

bottom of page