Recurrent retroprosthetic membranes in Boston Keratoprosthesis Type I: Incidence, risk factors, and complications

Abstract

Purpose

To identify pre-operative surgical factors associated with recurrent retroprosthetic membranes (RPMs) after Boston Keratoprosthesis (KPro) implantation, and to compare outcomes of patients with recurrent RPMs to those who had single or no RPM formation.

Methods

A single-center, retrospective study of all patients who underwent Boston Type I KPro implantation between 2006 and 2017 was performed. Patients with less than 1 year follow-up after device implantation were excluded. Patients were divided into 3 groups – no RPM, single RPM, and recurrent RPM. Recurrent RPM was defined as ≥2 visually significant RPM during follow-up. Statistical analysis was performed to assess factors associated with RPM status. Primary outcome measures included best-corrected visual acuity at 1 year and the need for RPM-related surgeries.

Results

109 of 148 KPro patients met the inclusion criteria. 45 % of patients had no RPM, 36.7 % had a single RPM, and 18.3 % had recurrent RPMs. At 1-year follow-up, visual acuity was decreased in both the recurrent (LogMAR 1.6) and single RPM (LogMar 1.18) groups compared to eyes with no RPM (LogMar 0.94). Eyes with recurrent RPM also required a significantly higher proportion of RPM-related surgery (p < 0.001). Eyes without prior penetrating keratoplasty (p = 0.04), pre-operative aniridia (p = 0.01), and post-operative occurrence of retinal detachments (p < 0.001) and corneal melts (p < 0.001) were found to be statistically significantly associated with RPM status.

Conclusions

Retroprosthetic membranes, especially when recurrent, play an important role in the secondary complications, visual outcomes, and device retention after KPro implantation.