Search

Purpose: To characterize the clinical and histologic features of primary graft failure after Descemet's stripping and automated endothelial keratoplasty (DSAEK)., Design: Retrospective observational case series., Participants: Sixteen cases of DSAEK graft failure from 15 patients, all with detailed histologic examination of failed graft tissue., Methods: Hematoxylin-eosin, periodic acid-Schiff staining, and light microscopy were used to examine the failed DSAEK graft tissue from all patients., Main Outcome Measures: Examination of specimens for corneal endothelial cell viability and host-donor interface characteristics., Results: Clinical history revealed that 88% (14/16) of studied DSAEK grafts detached before failure, and pathologic examination found that 75% (12/16) of failed grafts had atrophic corneal endothelium. Examples of residual host Descemet's membrane in the graft site and improper donor trephination were also identified., Conclusions: Marked loss of the corneal endothelium is the prominent feature of primary DSAEK graft failure. Examples of surgical features, such as incomplete Descemet's stripping and residual full-thickness cornea with a DSAEK graft, are shown.

Purpose: Orbital pleomorphic lipoma has been rarely reported in the literature. Although floretlike cells are characteristic of pleomorphic lipoma, they are not pathognomonic. We reviewed cases of prolapsed orbital fat and exenteration specimens to determine the significance of presence of these cells in the diagnosis of orbital pleomorphic lipoma., Design: Retrospective interventional case series with clinicopathologic correlation., Participants: Seventy-two specimens of 45 patients with prolapsed orbital fat and 74 exenteration specimens as controls., Intervention: Histologic review of the specimens including light microscopy, Masson trichrome staining, immunostaining for S100, CD34, CD68, terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL) assay, and transmission electron microscopy and review of clinical records and analysis of the data with generalized estimation equation., Main Outcome Measure: Evidence of histologic abnormalities in histologic specimens and clinical and demographic data., Results: Floretlike cells were present in 31 of 72 (43%) specimens of prolapsed orbital fat and in 12 of 74 (16%) orbital exenterations. Fewer than 6 florets were present in twenty 40x high-power fields in 15 (48%), 6 to 10 in 9 (29%), and >10 in 7 (23%) specimens. The florets stained positive for CD34 but not for S100 or CD68. TUNEL assay revealed significant nuclear pyknosis, and transmission electron microscopy disclosed spindle-shaped cells with abundant rough endoplasmic reticulum and no basement membrane. The mean age of patients with prolapsed orbital fat with florets was 67 years (range, 52-86). Of 31 samples, 29 (94%) were from males. Of 30 samples, 29 (97%) were located in the superotemporal conjunctiva; only one was located in the lower lid. There was significant association between the presence of florets and location of the prolapsed orbital fat (P = 0.0013) and gender (P = 0.0015)., Conclusion: Floretlike cells may be present in in situ and prolapsed orbital fat as a degenerative process. What some have called "orbital pleomorphic lipoma" is in fact only age-related orbital fat prolapse.