Eventual Endothelial Failure After Initial Corneal Clearing After a Detached Endothelial Graft in Fuchs Dystrophy.

We read the article by Lee et al with much interest. They recently reported a case with spontaneous resolution of cornea edema accompanied by improvement of visual recovery after a failed Descemet membrane endothelial keratoplasty (DMEK) procedure with a detached graft.[1] In the article, the authors speculated that there had been repopulation of corneal endothelial cells, after the failed DMEK, that involved migration of the endothelial cells from the outer side of the Descemet stripping area. Borkar et al[2] have also previously reported spontaneous healing of corneal edema after simple stripping without keratoplasty for Fuchs dystrophy.

We propose that the corneal endothelial cells could have originated from the donor graft that had been implanted with an upside-down orientation, which partially attached to the recipient's corneal back surface on the edge of the graft. In the article, severe corneal edema in the central area immediately after surgery was reported; however, clarity of the edematous area improved 5 months after DMEK. This healing process has been reported by Dirisamer et al[3] as “reverse clearance pattern after DMEK.” Although the mechanism of reverse clearance was not described in depth, they reported that it was the result of an upside-down graft insertion.

We would like to emphasize 3 points. First, the possibility of an upside-down graft insertion should be included when referring to the study by Dirisamer et al.[3] Second, the authors should include the area of graft detachment and the condition of the detached graft. Scheimpflug images of the anterior chamber using optical coherence tomography of the anterior segment would be very helpful. Third, the authors should explain how they recognized the graft orientation during DMEK surgery. Did they use an S stamp[4] or check the Moutsouris sign?[5]

If the DMEK graft is inserted in an upside-down orientation, the graft will only barely attach to the host cornea, often causing primary graft failure. However, there have been some cases with spontaneous healing after several months. We speculate that the corneal endothelial cells might have migrated from the small area that was partially folded and properly attached on the corneal stromal back surface, in the limited area on the edge of the graft. It was therefore possible to avoid surgical intervention and expect spontaneous improvement of corneal edema several weeks after the presumably failed DMEK procedure, because of upside-down insertion of the graft.