Keratoprosthesis has been used in adults since the mid-twentieth century [1–4]. However, the different types, surgical techniques, and models all have their limitations. Not until the advent of the Boston type 1 device, in the twenty-first century, did the popularity of the KPro increase. This device has been widely described [5, 6]. The Boston KPro has a fixed front plate fitting on top of a donor corneal rim. It is secured into place with a fenestrated back plate and locking titanium ring. The entire device is then implanted into the host corneal bed in the traditional fashion of penetrating keratoplastic surgery. The surgeon must select the correct size of the back plate, the correct size of the outer donor tissue to be used, and the correct optical power.

While keratoprostheses are usually reserved for patients that have sustained multiple failures or are otherwise poor candidates for penetrating keratoplasties, the success in these cases is impressive. It is thought that the newer Boston KProI model [7] with its fenestrated back plate, usage of a bandage contact lens, and the long-standing use of vancomycin drops as prophylaxis, has influenced the more favorable outcomes in recent reviews.

Prior to implantation, there are many considerations. A complete review of the workup, surgical procedure, and follow-up has been previously described [8]. In brief, it is important to take a full comprehensive history of the ocular anomalies and previous surgeries. In infants it is crucial to utilize an axial length to attempt to predict ocular power for the KPro. Surgery involves placement of a Flieringa ring for stability, assembly of the KPro, possible lensectomy and vitrectomy, and placement of a Kontur bandage lens. Postoperative management includes vancomycin 14 mg/ml and a fluoroquinolone drop to be instilled twice daily in the perioperative period and decreased to daily for the indefinite future. Examinations under anesthesia must be scheduled for infants or in any cases where a complete and thorough exam cannot be performed in clinic.

Most surgeons agree that the prognosis for success is diminished in adult candidates with ocular surface and autoimmune disease. In early comparisons between PKs and KPros, the failure rate of PKs in high-risk patients can approach to 56 % at 3 years [9]. Ma and colleagues noted that repeat transplant failures tend to occur early on, in the initial visual rehabilitation period, rendering a patient with no improvement in vision despite significant efforts from both the patient and the surgeon [10]. However, in Aquavella et al. [11] patients receiving KPros were noted to achieve best acuity in an average of 60 days, rather than the 6–9 months often quoted in PK literature. Seventy-four percent of these cases achieved a vision of 20/400 or better with 48 % ranging from 20/200 to 20/25 vision. In another series, 56 % of patients achieved best acuity in only 30 days, with 54 % of all eyes achieving 20/200 or better vision [12].

Despite good visual outcomes ranging from 20/200 to 20/100 or better in greater than 50 % of people [13–17] with poor prognosis, there are still many complications associated with the Boston type I procedure.

The most common nonsurgical complications are retroprosthetic membranes in most studies. In Greiner’s study, 55 % of eyes presented with membranes. Though treated with capsulotomies, 12.5 % of eyes were refractory to laser intervention. These numbers are similar to those in other studies where surgical capsulotomies were performed [14].

Persistent epithelial defects, not always reported, were found in 38 % of eyes in the Aldave series [15], treated with lateral tarsorrhaphy to help promote surface health and keep contact lens in place.

Glaucoma, which clearly coexists in the same patient population as those receiving the KPro transplant, is arguably the most significant for its impact on vision and lack of visual rehabilitation. In most studies, the majority (57.5–76 %) [13, 15] of patients already carried a diagnosis of glaucoma. The diagnosis was made on the basis of using IOP-lowering drops, documented nerve abnormalities, loss of RNFL, or prior glaucoma drainage device insertion.

With the difficulty to monitor glaucoma easily, information is gathered from appearance of optic nerve, visual field defects, and tactile intraocular pressure. Many series describe high intraocular pressure in follow-up visits with many incidences of progression of the optic neuropathy. In Grenier’s series [13] as well as Zerbe series [14], 15 % of eyes needed laser or surgical intervention to treat uncontrolled IOP. While 18 % were noted to have high IOP in Aldave’s series [15], only 2 % or one eye necessitated a tube shunt placement.

As unfortunate as glaucoma can be, endophthalmitis is the most feared KPro postoperative complication. Studies quote 5.4–12.5 % incidence, which is much higher than the quoted incidence in other intraocular surgeries [18–20].

Causes here may be loss of integrity of sclera-implant interface [21], poor contact lens hygiene, failure to use vancomycin and fluoroquinolone antibiotic drop prophylaxis, and erosion of implanted glaucoma drainage devices [13].

The differential of endophthalmitis does include sterile vitritis, the incidence of which seems to be higher following KPro surgery when compared with other intraocular surgeries. Nouri noted that this may be secondary to immune-mediated antigen reactions to tissue necrosis and melt [22].