(a) Endothelial precipitates following DMEK for treatment of Fuchs endothelial dystrophy in a 71-year-old individual. (b) Precipitates (arrowheads) also extend toward the initially denuded area demonstrating migration of donor-derived endothelial cells across the perimeter of the graft (arrows) (From Hos et al. ; reprint with kind permission)
15.6 Hypothesis for Less Rejection in DMEK Than in DSAEK
Besides the immunological reasons speculated on above, the (obvious) difference between DMEK and DSAEK is the disposition of stroma at Descemet membrane in DSAEK. It has been hypothesized that stromal and to a lesser extent endothelial antigen is responsible for initiating the immune response that then may lead to higher rejection rates in DSAEK than in DMEK (see above). However novel experimental findings might give reason for another hypothesis for increased rejection rates in DSAEK than in DMEK. Slnoniecka and coworkers recently published a study that demonstrated that the injury of the corneal stroma induces the secretion of substance P (SP) by corneal keratocytes . Consecutively SP stimulation induces the secretion of interleukin 8, a pro-inflammatory chemokine. Together with the highly interesting finding of Paunicka et al. from 2015, in which the authors demonstrated that a substance P burst, deriving from nerve damage in the published experiments led to abolition of immune privilege, also on the contralateral eye, introducing the term “sympathetic loss of immune privilege” (SLIP) . In the context of posterior lamellar keratoplasty, localized nerve damage occurs through the corneal incisions which would induce an initial SP burst. In DSAEK other than in DMEK, the remnant stroma and keratocytes are traumatized at time point of dissection and may therefore induce another (potentially chronic) SP secretion that would then abolish immune privilege and may lead to endothelial rejection, though weaker in characteristics than those in PK. In DMEK however, an additional SP burst is missing due to the lack of keratocytes which would be the reason for almost no endothelial rejections. To test this hypothesis on the one hand, animal models as described by Khodadoust and Silverstein in the 1960s could be used, but on the other hand, SP levels could be measured in DSAEK and DMEK specimens in vitro as first steps toward understanding the role of SP and SLIP in posterior lamellar keratoplasty.
15.7 Prevention and Treatment of Rejection in DMEK and DSAEK
Several studies have demonstrated that long-term low-dose therapy with topical steroids is successful in preventing endothelial rejection in PK. In a prospective study with more than 400 eyes that received a “normal-risk” PK, a once daily topical 1 % prednisolone acetate was demonstrated to significantly reduce the risk of graft rejection in comparison to eyes in which topical steroids were terminated 6 months after transplantation . Another study further demonstrated a positive effect of long-term 0.1 % fluorometholone (up to 9 years after PK) regarding the risk of endothelial corneal rejection .
For posterior lamellar keratoplasty, no comparable prospective studies are available. In terms of DSAEK which demonstrates clinically less severe though similar rates of rejection compared to PK, long-term use of topical steroids, e.g., once daily or every other day, seems reasonable. In DMEK, with only occasional rejection rates, careful consideration of benefits and potential side effects should be performed. However from our experience, continuation or early restart of topical steroids (unpreserved prednisolone acetate or dexamethasone) is recommended once immune cells or presumed immune cells are visible at the endothelium in routine slit lamp examinations, even if other signs of rejection such as edema or Khodadoust line are lacking.
15.8 Graft Failure or Immune Reaction?
It is currently unclear whether late-onset graft failures, accompanied with ongoing loss of endothelial cells, is caused by either previously undetected diseases of the donor endothelium or by subclinical immune reactions, in other words subclinical graft rejection. In this context better and more accurate noninvasive methods are necessary that would enable functional analysis of the graft: (i) to reliably detect and characterize immune cells in close proximity to endothelial cells, (ii) to measure endothelial cell function, and (iii) to detect early morphological changes of the donor endothelium accompanied with immune reactions or degenerative diseases such as Fuchs dystrophy. To achieve this, imaging methods with higher resolution than specular microscopy, for example, confocal laser scanning microscopy (CLSM), optical coherence microscopy , and two-photon microscopy (2 PM) [23, 24], would be suitable. All three technologies, potentially in combination, would enable imaging the transplanted endothelium at resolutions of approximately 1 μm (Fig. 15.2a) together with the option to use autofluorescence detection of intrinsic fluorophores to characterize immune cells and to quantify their activation state in 2 PM (Fig. 15.2b). In this context the most obvious technique, since already available in the clinic, would be CLSM. CLSM could give information on the morphology of endothelial precipitates in DMEK and DSAEK and potentially allow for easier decision-making regarding onset, duration, and intensity of an immunosuppressive therapy. A recent study by Monnereau et al. in 2014 demonstrated early changes of endothelial cells as a prequel of graft rejection in DMEK; however no immune cells or precipitates were detected and analyzed .
(a) En-face optical coherence microscopy (OCM) of pig corneal endothelium allowing detection of hexagonal cellular borders and subcellular details. (b) Autofluorescence signal of different immune cells allows differentiation by means of two-photon microscopy (From Gehlsen et al. , Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with kind permission)
15.9 Dry Eye Disease
Dry eye disease (DED) belongs among the most common ophthalmic conditions worldwide. It is generally accepted that the dysfunction of the lacrimal functional unit  leads to inflammation of the ocular surface as one of the core pathomechanisms of DED and consecutively to blurred vision, burning sensation, redness, tired eyes, and other symptoms. One reason for onset or exacerbation of DED is corneal refractive surgery such as LASIK or PTK [27, 28] and cataract surgery . In particular following refractive surgery, the dissection of corneal nerves interrupts the corneal reflex arc, thereby reducing lacrimation and at the same time triggering an inflammatory response through the release of neuron-derived peptides such as substance P and CGRP . The same may be hypothesized following perforating keratoplasty because all corneal nerves are dissected during trephination. Interestingly until now, only little information on DED following PK has been available, possibly because of consecutive anesthesia of the cornea and therefore lack of subjective symptoms. One study by Kuechle et al. demonstrated that dry eye disease is a risk factor for graft rejection in PK . However, it is still unclear to what extent DED develops following PK.
Regarding DSAEK and DMEK, no information on DED either before or after surgery is available. Recently we demonstrated by means of CLSM that corneal incisions during DMEK surgery lead to a significant alteration of the density and function of the subbasal nerve plexus of the cornea  (Fig. 15.3). Although all parameters tested (including corneal sensitivity) recovered within 4–10 months, it is likely that during this recovery period or beyond, DED may be present and like in PK might have an influence on subclinical or clinical graft rejection.
Corneal nerve alterations following DMEK. (a–d) Anterior segment photographies. (e–h) Anterior segement optical coherence tomography and (i–l) CLSM imaging of subbasal nerve plexus demonstrates reduction of nerve length, number, and branches at 1 week after surgery which fully recover at 4 and 10 months post surgery (From Bucher et al. , reprint with kind permission)
Overall prospective clinical trials are needed to investigate the role of preexisting DED in the context of corneal transplantation and in particular its influence on graft survival.
Regarding immune reactions following corneal transplantation, DMEK surgery demonstrates advantages through very limited rejection episodes in comparison to DSAEK, which shows almost similar rejection rates as PK. Several explanations and hypotheses exist that could explain reduced rejection rates in DMEK and less severe immune responses in DSAEK that imply potential benefit of prolonged topical steroids. In the future, additional basic immunological and clinical studies together with novel imaging technologies will enable earlier and more accurate diagnosis of graft rejection. These approaches will allow differentiation of nonimmunological graft failures and more appropriate application of immunosuppressive therapy. The role of DED in the context of posterior lamellar keratoplasty needs to be elucidated and might have an influence on graft survival as demonstrated for PK.
Melles GR, Ong TS, Ververs B, van der Wees J. Descemet membrane endothelial keratoplasty (DMEK). Cornea. 2006;25(8):987–90. doi:10.1097/01.ico.0000248385.16896.34 [published Online First: Epub Date].CrossRefPubMed