Endothelial keratoplasty (EK) has replaced PK as the treatment of choice for corneal edema arising from corneal endothelial disorders such as Fuchs endothelial dystrophy and pseudophakic bullous keratopathy, with DSAEK and Descemet’s membrane endothelial keratoplasty (DMEK) representing the most commonly performed EK techniques. In DSAEK, the host DM and endothelium are replaced by donor posterior stroma, DM, and endothelium, whereas in DMEK the graft consists only of donor DM and endothelium. Nevertheless, the presence of marked anterior corneal scarring may limit the final visual outcome; hence, PK has traditionally been the surgical treatment of choice in the presence of corneal fibrosis. Alternative therapeutic options have been reported in the past, including a two-step procedure, such as DSAEK combined with transepithelial PTK and adjunctive mitomycin C [5, 6].
Using in vivo laser confocal microscopy, Morishige et al. found that preoperative SEF observed in bullous keratopathy was usually detected in patients with long-standing corneal stromal edema of more than 12 months [7]. In addition, in approximately one-third of the patients who underwent DSAEK, persistent SEF was still detected 1 year postoperatively. Taking into consideration this observation, they hypothesized that the duration of preoperative stromal edema is a key determining factor for the resolution of the SEF and the final DSAEK outcome [7]. Nevertheless, in our case, we observed complete resolution of the severe anterior corneal scarring 12 months following surgery, despite the long-standing bullous keratopathy.
The development and persistence of mature myofibroblasts in the corneal stroma are key steps in the development of stromal fibrosis [8]. At the site of epithelial basement membrane (EBM) injury, transforming growth factor (TGF)-J3 and platelet-derived growth factor released from epithelial cells penetrate the stroma and initiate the development of myofibroblasts [8], which contribute to the abnormal distribution of extracellular matrix (ECM) [9, 10]. Similarly to EBM injury, it has been reported that corneal myofibroblasts and stromal fibrosis persisted when the Descemet membrane was damaged [11]. In addition, activation of TGF-J3 signaling in Fuchs endothelial dystrophy mediates excessive accumulation of ECM proteins [12]. In long-standing corneal edema with epithelial bullae due to endothelial dysfunction, both the EBM and Descemet membrane are defective. Thus, corneal basement membranes (EBM and Descemet membrane) are the modulators of the corneal fibrosis response [8]. Endothelial keratoplasty restores adequate hydration of the stroma by restoring Descemet membrane function.
In this regard, a possible mechanism explaining regression of SEF in our case could be that the resolution of the stromal edema observed following DSAEK induced the clearance of myofibroblasts, the repopulation of keratocytes and reabsorption of the abnormal ECM components. We also performed epithelial removal during our patient’s procedure. Epithelial debridement has been shown to induce senescence of stromal fibroblasts and may thus be involved in the resolution of corneal fibrosis [13]. In addition, the use of topical steroids during the postoperative period may have played a key role, as its use has been shown to reduce scar tissue formation [14].
Corneal fibrosis may undergo spontaneous resolution over months to years with a return of partial to full corneal transparency [8, 15]. Thus, we claim that PK should not be systematically proposed and DSAEK could be considered as a therapeutic option in patients with endothelial dysfunction and stromal fibrosis, especially in cases of patients with low visual potential or who cannot undergo PK, such as patients under anticoagulant therapy.