POSTOPERATIVE POSTERIOR SEGMENT COMPLICATIONS AFTER BOSTON TYPE 1 KERATOPROSTHESIS: Incidence, Risk Factors, and Intermediate-Term Outcomes

Risk factors for early and late retinal detachment after boston type I keratoprosthesis surgery

PURPOSE

To determine the differential risk factors for retinal detachment (RD) after Boston Type 1 Keratoprosthesis (B1KPro) during the first year after surgery (early RD; ERD) and afterwards (late RD; LRD).

METHODS

Retrospective cohort study of 94 eyes (90 patients) undergoing B1KPro implantation at Centro de Oftalmología Barraquer from June 2006 to July 2022 with a minimum follow-up of one year.

RESULTS

The incidence of RD in the whole sample after B1KPro implantation was 29% (27/94), occurring in 21% of eyes (ERD; 20/94) the first year, and in 7% afterwards (LRD; 7/94). Hypotony was revealed as a significant risk factor for ERD (16/19 = 84%; OR = 14.5, P < 0.0001), together with aphakia (16/20 = 80%; OR = 5.9, P = 0.004), intraocular lens (IOL) removal (4/8 = 50%; OR = 6.2, P = 0.03), previous choroidal detachment (CD; 7/20 = 35%; OR = 35.5, P = 0.001) and previous pars plana vitrectomy (PPV; 15/20 = 75%; OR = 6.6, P = 0.006). Multivariate analysis included hypotony and PPV at any time as variables significantly associated with ERD occurrence. The occurrence of LRD was significantly related to peripheral retinal changes, detected few days before (4/6 = 66%; OR = 65, P < 0.0001).

CONCLUSION

Hypotony emerged as one of the main factors influencing ERD occurrence after B1KPro implantation. Additionally, previous CD, PPV, aphakia, and IOL removal should be considered to estimate the risk of postoperative retinal detachment. In contrast, peripheral retinal changes after B1KPro seem to be significantly related to LRD occurrence. This study reinforces the need for frequent intraocular pressure evaluation, and comprehensive periodic retinal assessment over time.

KEY MESSAGES

What is known Retinal detachment (RD) is a severe complication after Boston Type 1 Keratoprosthesis (B1KPro) implantation, resulting in a permanent reduction of visual acuity in a clinically relevant percentage of eyes. The time of RD occurrence after B1KPro implantation differs in the literature but is reported to be more common within the first year. Previous aphakia, choroidal detachment, or RD repair, intraocular lens removal, vitritis, or history of Nd-YAG laser have been described as risk factors for RD after B1KPro surgery in samples with varying follow-ups.

WHAT IS NEW

Multivariate analysis in the first year after B1KPro surgery showed hypotony and pars plana vitrectomy at any time as significant risk factors for early RD. Late RD cases, occurring after one year following the B1KPro surgery, seem to be significantly related to peripheral retinal changes suffered a few days before loss of visual acuity and RD diagnosis. It is plausible to think that risk factors for RD after B1KPro differ in short and long postoperative times.

Ultra-wide field imaging to assess the optic nerve and retina in Boston type I and II keratoprosthesis patients

Background

The ability to view the posterior segment in keratoprosthesis (Kpro) implanted patients is limited. The purpose of this retrospective, observational study was to investigate the use of ultra-wide field (UWF) scanning laser ophthalmoscopy imaging and its utility for serial evaluation of the retina and optic nerve in patients with either a Boston type I or II Kpro.

Methods

A retrospective chart review was performed for patients with a Boston type I or II Kpro seen at The Ohio State University Wexner Medical Center. Images were graded for quality by two masked observers on a defined four-point scale (“Poor”, “Fair”, “Good”, or “Very good”) and assessed for visible posterior segment anatomy. Interobserver agreement was described using the Kappa statistic coefficient (κ) with 95% confidence intervals.

Results

A total of 19 eyes from 17 patients were included in this study. Eighteen eyes had a type I Kpro, while one eye had a type II Kpro. UWF imaging from 41 patient visits were reviewed by two observers. Interobserver agreement between the two graders was fair for image quality (κ = 0.36), moderate for visibility of the macula with discernible details (κ = 0.59), moderate for visibility of the anterior retina with discernable details (κ = 0.60), and perfect agreement for visibility of the optic nerve with discernible details (κ = 1.0). In 6 eyes, UWF imaging was performed longitudinally (range 3–9 individual visits), allowing for long-term follow-up (range 3–46 months) of posterior segment clinical pathology.

Conclusions

UWF imaging provides adequate and reliable visualization of the posterior segment in Kpro implanted patients. This imaging modality allowed for noninvasive longitudinal monitoring of retinal and optic nerve disease in this selected patient population.

The treatment of end-stage corneal disease: penetrating keratoplasty compared with Boston type 1 keratoprosthesis

Penetrating keratoplasty (PKP) yields excellent results for restoring vision in end-stage corneal diseases. However, its success is limited to high-risk diseases such as aniridia, chemical burns, autoimmune corneal diseases, and herpetic eye disease. Boston type 1 keratoprosthesis (BKPro) offers another option to these patients. Since 1992, improvements in perioperative management and device construction have significantly increased the use of BKPro worldwide and challenged the therapeutic role of PKP in these patients. This review aims to evaluate BKPro's place in the treatment algorithm of these high-risk patients to assist surgeons' decision-making. PKP and BKPro are compared in three outcome categories: visual acuity, graft retention and failure, and complications profile. Special attention is given to comparing secondary BKPro versus repeated PKP as well as primary BKPro versus primary PKP. We conclude that secondary BKPro bears a better prognosis than repeated PKP in most high-risk patients. Similarly, primary BKPro likely confers improved outcomes over primary PKP in most high-risk recipients. However, current evidence is based on retrospective designs, and controlled prospective randomized trials are required to validate these conclusions.

Retinal Detachment in Eyes with Boston Type 1 Keratoprosthesis: Surgical Techniques and Mid-Term Outcomes

PURPOSE

To evaluate the mid-term outcomes of pars plana vitrectomy (PPV) performed for retinal detachment (RD) repair after Boston type 1 keratoprosthesis (KPro) implantation.

METHODS

Retrospective chart review of KPro implanted at the Stein Eye Institute presenting with RD and treated by PPV. Functional success was defined as a postoperative visual acuity (VA) maintained within 2 Snellen lines of the CDVA measured before the development of the RD (baseline) and anatomical success as an attached retina after the PPV. Kaplan-Meyer survival analyses were performed.

RESULTS

Among the 224 KPro performed, 28 (15.2%) RD were identified, of which 21 (9.4%) were included. The mean follow-up was 42.5 ± 27.3 months. Vitreoretinal proliferation was present in 18/21 (85.7%) eyes. Surgical techniques were adapted to the complex anterior segment anatomy of KPro eyes. Anatomical success was achieved in 18/21 (85.7%) eyes. Functional success occurred in 17/21 (81.0%) eyes, and 5/21 (23.8%) eyes reached 20/400 or better VA at final follow-up. The KPro was retained in 11/21 (52.4%) eyes. The retention rate decreased from 94.7% at 1 year to 53.5% at 5 years. The most frequent complications were retroprosthetic membrane (47.6%) and corneal melt (23.8%).

CONCLUSION

Modified PPV techniques resulted in relatively good mid-term anatomic, functional, and retention rate outcomes, given the severity of RD at presentation, and the numerous preoperative comorbidities of KPro eyes.