Effects of Systemic Diseases on Graft Preparation in Descemet Membrane Endothelial Keratoplasty

Supplemental Digital Content is Available in the Text. Objectives: To evaluate the effects of the systemic diseases and drugs of the donor on Descemet membrane (DM) graft preparation. Methods: Seventy-eight corneas of 58 donors, of whom the DM grafts were used in Descemet membrane endothelial keratoplasty (DMEK) surgery, between January 2018 and January 2020, were enrolled in this retrospective study. The hospital records of the donors were analyzed. Age, sex, blood type, systemic diseases, and drugs; complete blood count; biochemistry panel for liver and kidney functions in the past 48 hours; and the drugs used in the hospital, if any, in the past 24 hours were recorded. The grafts with tears that occurred while preparation were included in group 1, and the successful grafts with no tears were included in group 2. Results: There were no statistically significant differences in the characteristics of the donors between groups. However, breast cancer and the use of sevelamer were found to be significantly higher in group 1 (P=0.010, P=0.033, respectively). No statistically significant difference in the use of other drugs was found between groups. Conclusion: Although diabetic donors have been reported to be inappropriate candidates for the preparation of DM grafts for DMEK, most of the donors with several systemic diseases including diabetes can be used in DMEK surgery, with the right technique in DM graft preparation.

D escemet membrane endothelial keratoplasty (DMEK) is a lamellar keratoplasty, in which the posterior corneal layer is transplanted selectively with a lower rejection risk, a faster visual recovery, and a better anatomical restoration than penetrating keratoplasty. 1,2 The procedure has been modified since it was first described in 2006. [1][2][3] Surgically induced astigmatism and suturerelated problems are minimal because of the minimal manipulation on the host anterior surface. 4 Several techniques have been described in the preparation of Descemet membrane (DM) graft in DMEK. [5][6][7] Successful DM graft preparation without any tears is a critical step in DMEK surgery. It has been shown that the surgeon's skill and experience are the most important factors in dissection techniques and rescuing the central and peripheral tears in grafts with tight DM adhesions. 8 These tears in the DM graft complicate the effective use of this delicate tissue of 10 mm in thickness. Apart from the surgical manipulation, the structural differences in donor tissue also have a role in the tears in DM grafts while preparing the tissue. 9 Moreover, the effect of the systemic disease and the drugs of the donor in a successful DM graft remain unclear.
In diabetic patients, abnormal adhesions between DM and posterior stroma and an increased risk of tissue preparation failure have been reported. 10,11 These findings made us think that systemic drugs and other systemic diseases may have an effect on graft preparation. This study aims to evaluate the effect of systemic diseases and drugs of the donor on DM graft preparation.

MATERIALS AND METHODS
Seventy-eight corneas of 58 donors, DM grafts from whom were used in DMEK surgery in an ophthalmology clinic of a training and research hospital between January 2018 and January 2020, were enrolled in this retrospective study. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards (Adana City Training and Research Hospital, January 29,2020, 49, 701).
The grafts from donors were harvested by an experienced personnel of our eye bank. The hospital records of the donors and the donor information form were analyzed (Table 1). Age, sex, blood type, systemic diseases, and drugs; complete blood count; biochemistry panel for liver and kidney functions in the past 48 hours; and the drugs used in the hospital, if any, in the past 24 hours were recorded. The donors were from White race.
The video records of DM graft preparation were analyzed. All the grafts were prepared just before the procedure in the operating room by the same experienced surgeon (Y.K.). The grafts were divided into two groups. The grafts with tears occurred during preparation were included in group 1, and the successful grafts with no tears were included in group 2. The surgical technique is shared in Videos 1 and 2, Supplemental Digital Content, http:// links.lww.com/ICL/A237, http://links.lww.com/ICL/A238 (DM graft with tear and without tear, respectively). The grafts from donors with a history of ocular surgery, Descemet folds, and/or endothelial precipitates were not used for the DM graft. 12 The donor corneas were preserved in Eusol-C (Corneal Chamber, Alchimia, Ponte San Nicolo, Italy) at +4°C. Descemet membrane graft preparation and DMEK surgery were performed within 7 days.
The first 200 cases of the surgeon, that is, the surgeries that include another procedure, such as synechiotomy, vitrectomy, and anterior-chamber lens implantation were excluded from the study.

Graft Preparation
The diameter of the graft was determined according to the diameter of the cornea from limbus to limbus of the recipient's eye. All grafts were prepared with the guidance of a technique described as submerged corneas using backgrounds away (SCUBA). 13 The donor cornea was placed endothelial side up on a vacuum punch (Katena Products, Inc., New Jersey, USA). Trypan blue 0.06% solution was used to stain the edges of DM. Under the fluid, peripheral 180°of DM was stripped, from the trabecular meshwork to the central cornea by using a crescent knife, trying not to form a tear. The Descemet membrane was grasped with a Sinskey hook and 360°of stripping was performed using tying forceps. When nearly half of the DM was stripped, the graft was restained. Then, superficial partial-thickness trephination was performed from the endothelial side with a diameter that was determined before (7.50-8.50 mm). Then, the DM was grasped again with a pair of tying forceps and stripped until 360°of DM was free from the stroma. The donor graft was restained by trypan blue 0.06% and  In case of a peripheral DM tear occurrence, while stripping from the trabecular side, the same procedure was repeated from the opposite side. If again a tear was formed, then central punching was performed by not including the tear (Fig. 2). These grafts were included in group 1, and the grafts without any tear formation were included in group 2.

Statistical Analysis
Data analysis was performed using the Statistical Package for Social Sciences for Windows software (SPSS version 16.0, SPSS Inc., Chicago, IL). The normality distribution of the variables was tested. The descriptive statistics of the normally distributed continuous variables were expressed as mean 6 SD. The normally distributed variables were compared between the groups using a Student t-test. Categorical variables were presented as frequency (%) and compared between the groups using the chi-square test or Fisher exact test. According to the data we used in the study, the power analysis of the sample size was found to be 0.82 in the preoperative analysis. Differences with a P value of less than 0.05 were considered statistically significant.
The use of systemic drugs is summarized in Table 3. Sevelamer use was only found in group 1 (n¼2, 8.7%), and the difference was statistically significant between groups (P¼0.030). When we looked at the drugs included in our study, no statistically significant difference in the use of other drugs was found between groups (Table 3).
There was no statistically significant difference between groups in the use of drugs in the past 48 hours for respiratory or cardiac arrest, such as dopamine, amiodarone, and adrenaline (P¼0.130). The increased levels above the normal range in blood urea nitrogen (BUN) and creatinine were found to be higher in group 1 than group 2 (87.5% vs. 66%, P¼0.050). No statistically significant difference was found in increased levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) in donors when compared between groups (P¼0.984). In addition, there was no statistically significant difference in lower hemoglobin levels between groups (P¼0.701).  CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease.

DISCUSSION
The first step of DMEK surgery is DM graft preparation. 9 One of the most concerning reasons in graft preparation is graft loss, whereas recent studies have reported successful outcomes with partially implanted grafts. [13][14][15] The most common cause of graft loss is the tears within the graft due to the tight adhesions or increased fragility of the graft. 16 Although the procedure seems to be standardized, the peripheral tears in the DM graft remain a potential risk. Although small tears may be salvageable, an extra graft must be reserved in case of a large tear. 10 The differences in histochemical properties of the DM graft and the thickness of the graft may cause a change in the resistance at the peeling stage of the preparation step. The systemic diseases and drugs of the donor may also affect the elasticity and resistance of the tissue. The structural stress of the tissue increases due to the differences in adhesions between tissues. There is a narrow transition zone that includes an amorphous extracellular matrix, called interfacial matrix, between the DM and the stroma. Extracellular matrix and adhesive proteins, such as fibronectin, vitronectin, amyloid P, osteonectin/SPARC, fibrillin-1, fibulin-1 ve fibulin-3, keratoepithelin, and collagen type VI and VIII, are found immunohistochemically in this zone. 17,18 Descemet membrane graft tears during peeling may be a result of these adhesive properties of the interfacial matrix. During DM graft preparation, adhesions between the posterior stroma and the DM may resolve spontaneously or result in DM tears if the tractional forces exceed the tensile adhesive strength. 19 The structural and biochemical differences between donors may cause the different adhesive forces between the posterior stroma and the DM. 18 Failure in DM graft preparation has been reported between 4.2% and 6.7%. 8,[20][21][22] The most commonly used technique in graft preparation is the SCUBA technique, in which the success rate has been reported over 95%. 21,23 In this study, only the SCUBA technique is used to prepare the DM grafts. As we know that with younger donors, the adhesive forces and the possibility of tears are higher, donors younger than 40 years were also not included. 24 No agerelated or sex-related factor was found to affect the failure rate of graft preparation in this study.
As far as we know, no investigation between blood type and DM tears was made. In this study, the highest rate of DM tears was found in donors with a blood type of A Rh+, but no statistically significant difference was found. This finding may be due to the higher rates of A Rh+ blood type in our population.
Previous studies have shown that diabetes, hyperlipidemia, obesity, and chronic renal failure in the donors cause failure in preparation of the DM graft. 20,25 It has been suggested that in diabetes, the tight adhesions between the posterior stroma and the DM may be the cause of DM tears. 25 In our study, no statistically significant difference was found in the rate of diabetes between groups. All the DM grafts from diabetic donors, whether intact or with tears, were used in the surgery and no complications occurred. In addition, the use of insulin and oral antidiabetic in donors had no significant effect on DM tears, although there seems to be a trend toward a statistically significant higher tear rate with insulin use (Table 3). In a study by Tien-en Tan et al., 26 it has been shown 4.7% of failure in DM graft preparation and no effect of diabetes on DM tears. Thus, it is believed that not only metabolic diseases, such as diabetes, but also multiple structural differences may be related to the occurrence of DM tears while graft preparation. 26 It has been reported that systemic diseases, cause of death, and age of the donor may have a role in the failure of DM graft preparation. 20,27 In our study, among systemic diseases of donors, only the rate of breast cancer was found to be significantly higher in group 1. Increased inflammatory response, vascular endothelial dysfunction, increased vascular endothelial growth factor response, and increased oxidative stress are parts of the pathophysiological process of breast cancer. 28 Thus, the increased oxidative stress may be the cause of the dysfunction of the DM-endothelium complex of the graft and leads to the failure of DM graft preparation. Prospective studies might be required to further approve this finding.
When systemic drugs that were used more than 1 year were analyzed, only sevelamer use was found to be more common in group 1. Sevelamer is a phosphate binding agent indicated for the control of hyperphosphatemia, it effects the chloride ions by exchanging and not absorbed from gastrointestinal tract. 29 Therefore, the effects on the donor may not have made a significant difference. However, it is a limiting situation that there are only 2 donors in group 1 who use sevelamer and there are no donors in group 2 who use sevelamer. With the prospective studies, these findings might be investigated.
Although no statistically significant differences were found in anemia or liver function tests between groups, renal function tests (BUN or creatinine) were found to be more common in group 1. The level of urea in the aqueous humor has been shown to increase with increased urea levels in the serum. 30 The design of this study is not suitable to understand the effects of increased blood urea and creatinine levels on the DM-endothelium complex, so further studies are needed to reveal the pathophysiology.
The limitations of our study are its retrospective nature and the lack of a control group that includes naive donors without any systemic disease or use of any drug. However, it is almost impossible to find enough healthy donors in relatively old age within the criteria of donor bank to form a control group. There may be other drugs off our list that can affect the DM tear rate, so additional studies are needed (Table 3).
In conclusion, most of the death in elderly occurs with individuals with systemic diseases. On the other hand, the need