A and B, Hematoxylin-eosin staining of a thick melt-associated retroprosthetic membrane showing diffuse cellular lymphocytic inflammation (yellow arrowhead) and stratified cellular proliferation on the anterior surface suggestive of epithelial ingrowth (black arrowhead). Fragments of Descemet’s membrane are seen within the membrane (white arrowhead). C and D, Positive staining for pankeratin (cytokeratin AE1/3) (C) and cytokeratin 7 (D) is consistent with epithelial ingrowth. E, Vimentin staining was absent in epithelium. F and G, Significant vascularity was noted with CD34 (F) and α–smooth muscle actin (G) staining of the vascular endothelium and smooth muscle, respectively.
A and B, Fluorescence in situ hybridization using a dual color X/Y (spectrum orange/spectrum green) probe and diamidino-2-phenylindole counterstaining showing representative interphase cells from donor-recipient sex-mismatched retroprosthetic membranes. C, In retroprosthetic membranes obtained from 2 female recipients who received male corneas, 100% of the cells showed an XX-karyotype. D, In one retroprosthetic membrane from a male recipient who received a female cornea, 100% of the cells showed an XY-karyotype. In another retroprosthetic membrane from a male recipient who received a female cornea, a mixture of XY and XX cells was observed.
Hou JH, Sivaraman KR, de la Cruz J, Lin AY, Cortina MS. Histopathological and Immunohistochemical Analysis of Melt-Associated Retroprosthetic Membranes in the Boston Type 1 Keratoprosthesis. JAMA Ophthalmol. 2014;132(9):1133-1136. doi:10.1001/jamaophthalmol.2014.1959
Retroprosthetic membrane (RPM) formation is the most common complication associated with the Boston type 1 keratoprosthesis and has been associated with corneal melt.
To identify the histological and immunohistochemical characteristics of RPMs associated with corneal melt.
Design, Setting, and Participants
Observational histopathological case series at a tertiary eye care referral center among patients who underwent Boston type 1 keratoprosthesis explantation because of donor corneal melt at the Illinois Eye and Ear Infirmary between January 1, 2011, and January 1, 2012.
Seven RPM specimens from 7 eyes were stained with hematoxylin-eosin, cytokeratin 7, cytokeratin AE1/3, smooth muscle actin, vimentin, and CD34. Light microscopy was used to evaluate specimens for inflammation and epithelial ingrowth. XY-karyotyping using fluorescence in situ hybridization was performed on 4 specimens with known donor-recipient sex mismatch.
Main Outcomes and Measures
Histological and immunohistochemical characteristics of RPMs.
Inflammatory cells were present in 4 of 7 RPMs. In 3 of 4 sex-mismatched specimens, tissue XY-karyotyping of the RPM interphase cells was consistent with the host sex karyotype. The fourth specimen showed a mixture of recipient-type and donor-type cells.
Conclusions and Relevance
Melt-associated RPMs show variable degrees of inflammation. Most membranes seem to originate from a proliferation of host cells, but donor tissue may contribute in some cases.
The Boston type 1 keratoprosthesis (KPro) is an artificial cornea used for the management of corneal diseases not amenable to standard keratoplasty.1 Despite improved outcomes, the Boston KPro continues to carry significant risk for complications. Of these, the most common is retroprosthetic membrane (RPM) formation, which occurs in 25% to 65% of eyes.2- 7
Despite their high incidence, RPMs are poorly understood. A histological analysis of 4 RPM specimens by Stacy et al8 showed minimal inflammation in the membranes. However, other investigators have suggested that anterior chamber inflammation is a risk factor for RPM formation and that the use of more biocompatible materials (eg, titanium) in the Boston KPro construct or the administration of perioperative corticosteroids may retard RPM growth.2,9- 11
Given recent data that RPM formation may increase the risk of donor corneal melt,12 further investigation of RPMs is warranted. To the best of our knowledge, no dedicated histological study of melt-associated RPMs has previously been performed. To clarify the role of inflammation in RPMs and to further characterize melt-associated RPMs, we performed a histopathological and immunohistochemical analysis of RPM specimens obtained from Boston KPro devices explanted because of donor corneal melt.
After obtaining institutional review board approval at the University of Illinois at Chicago, a retrospective evaluation of RPM specimens obtained from Boston KPros explanted because of corneal melt was performed. All available RPM pathology specimens obtained between January 1, 2011, and January 1, 2012, at our institution were identified (institutional review board exemption for informed consent was obtained). The clinical history for each patient was reviewed.
Immunohistochemistry was used to characterize the membranes. Anti-CD34 (clone QBEnd/10; Ventana Medical Systems) and α–smooth muscle actin (SMA) stains (clone 1A4; Ventana Medical Systems) were used to distinguish normal stromal keratocytes (CD34 positive) from fibrous keratocytic downgrowth (SMA positive) and to highlight vascularity within the membranes.13 Pankeratin (clone AE1/AE3/PCK36; Ventana Medical Systems) was used to identify epithelium. Anti–cytokeratin 7 (clone OV-TL 12/30; Ventana Medical Systems) was used to distinguish conjunctival epithelium from corneal epithelium.14 Vimentin (clone V9; Ventana Medical Systems) was used as a negative control for identifying epithelium. All immunohistochemical staining was performed on an automated immunostainer (Benchmark XT; Ventana Medical Systems) and was reviewed by one of us who is a board-certified ocular pathologist (A.Y.L.).
To further characterize the cellular origin (donor or recipient) of the RPM tissue, fluorescence in situ hybridization (FISH) was performed on a subset of specimens. Donor sex and recipient sex for each study patient were reviewed for donor-recipient sex mismatch. Cases in which a male patient received a female donor cornea or a female patient received a male donor cornea were identified.
Fluorescence in situ hybridization was then performed using a dual color X/Y (alpha satellite, spectrum orange/satellite III, and spectrum green) centromere enumeration probe (Vysis).15 Results were then scored by a cytogenetics consultant masked to the sex of the recipients and donors to determine the XY-karyotype of the interphase cellular population.
In total, 7 RPM specimens obtained from 7 consecutive eyes following Boston KPro explantation because of donor corneal melt were included in the study. Demographic data for each patient are summarized in the Table.
A lymphocytic infiltrate was identified on hematoxylin-eosin staining in 4 of 7 samples (Figure 1). Two had mild inflammation, while 2 others had severe inflammation. Both specimens with severe inflammation were obtained from patients treated for aniridia (Table).
A proliferation of nonkeratinizing stratified epithelium was present along the surface of the explanted membrane specimens in 4 of 7 samples (Figure 1). Pankeratin (cytokeratin AE1/3) and cytokeratin 7 showed strong cytoplasmic immunoreactivity suggestive of epithelial ingrowth of conjunctival origin (Figure 1). Epithelial cells were negative for vimentin.
Of 4 patients with histological evidence of epithelial ingrowth, one (case 2 in the Table) had clinical epithelial downgrowth at the time of Boston KPro explantation, with cyclitic membranes, hypotony, and a dense epiretinal membrane. In 2 specimens with no epithelial ingrowth, epithelium was found within stumps of tissue removed from the backplate holes.
In total, 5 of 7 specimens had SMA staining consistent with fibrous stromal (keratocytic) downgrowth. Six of them had significant vascularity on CD34 and SMA staining.
In total, 4 cases of known donor-recipient sex mismatch were identified (Table). In 3 of 4 cases, 100% of the interphase cells in the RPM specimen showed an XY-karyotype consistent with the recipient’s sex (Figure 2). In the fourth case, which involved a male recipient and a female donor cornea (case 6 in the Table), 69% of the interphase cells were XY, and 31% of the cells were XX (Figure 2). The findings were consistent with mixed contributions to the RPM from both recipient and donor tissues.
To our knowledge, this is the first dedicated histological study of RPMs explanted from eyes with a Boston type 1 KPro and sterile corneal melt. The only other histological study of RPMs, by Stacy et al,8 was based on 4 RPMs (2 were associated with corneal melt). The authors concluded that inflammation was not a likely requisite for the formation of RPMs.8 In contrast, our study showed chronic inflammatory cells in 4 of 7 RPMs. Therefore, it is likely that inflammation has a role in at least some cases of melt-associated RPMs. Furthermore, therapeutic intervention with local or systemic corticosteroids may benefit a subset of patients with inflammatory RPM components (eg, those with aniridia in our study).
Based on observations of stromal downgrowth through breaks in Descemet membrane at the wound margin, Stacy et al8 proposed that RPMs are derived from host corneal stroma. However, our FISH analysis of sex-mismatched donor-host pairs suggested that, while most RPMs seem to originate from host tissue, donor tissue may also contribute in some eyes. Based on this, we postulate that stromal downgrowth from donor tissue may occur as well at the wound margin or through the backplate holes.
Another notable finding in this study was the presence of epithelial ingrowth between the backplate and RPM in 4 of 7 of our melt-associated specimens. The isolated presence of epithelium in the backplate holes in 2 of our RPMs suggests that epithelium may localize to the posterior backplate through the holes. Despite the high incidence of epithelial ingrowth between the backplate and RPM in our study (in 4 of 7 cases), clinical evidence of epithelial downgrowth was rare. One explanation for this is that the RPM may provide a physical barrier to invasion of epithelium into the anterior chamber. This conjecture is consistent with the low rates of aqueous leakage on Seidel testing in our patients despite the presence of full-thickness corneal melts.
The major weaknesses of the study are its retrospective nature and the small sample size, which may limit the extrapolation of these findings to all melt-associated RPMs. Because tissue was analyzed retrospectively, the quality of tissue obtained varied.
Melt-associated RPMs seem to be heterogeneous tissues showing varying degrees of inflammation. Retroprosthetic membranes appear to be derived mainly from host cells; however, contributions from donor tissue seem possible. Further study is needed to better describe the inciting factors for RPM formation, the role of inflammation, and the relative contributions of host and donor corneal tissues.
Submitted for Publication: October 21, 2013; final revision received April 8, 2014; accepted April 11, 2014.
Corresponding Author: Maria Soledad Cortina, MD, Department of Ophthalmology and Visual Sciences, University of Illinois at Chicago, 1855 W Taylor St, Mail Code 648, Chicago, IL 60612 (email@example.com).
Published Online: July 10, 2014. doi:10.1001/jamaophthalmol.2014.1959.
Author Contributions: Dr Cortina had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Hou, Sivaraman, de la Cruz.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: de la Cruz.
Obtained funding: Hou.
Administrative, technical, or material support: Hou, Sivaraman, Lin.
Study supervision: de la Cruz, Lin, Cortina.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by an unrestricted research grant from the Illinois Society for the Prevention of Blindness.
Role of the Sponsor: The funding source was not involved in any aspect of study design or conduct; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Previous Presentation: Preliminary data from this study were presented at the XXX Congress of the European Society of Cataract and Refractive Surgeons; September 10, 2012; Milan, Italy.
Additional Contributions: Amy Wu, BS, and Valerie Lindgren, PhD, of the Department of Pathology, University of Illinois at Chicago, assisted with the fluorescence in situ hybridization studies. There was no financial compensation for their assistance.