Zachary M. Mayko

The First 100 Eyes of Standardized Descemet Stripping Automated Endothelial Keratoplasty versus Standardized Descemet Membrane Endothelial Keratoplasty.

PURPOSE To compare results of the first 100 eyes of Descemet stripping automated endothelial keratoplasty (DSAEK) and Descemet membrane endothelial keratoplasty (DMEK) that were performed with a standardized technique at a single institution. DESIGN Single-center, retrospective case series. PARTICIPANTS The first 100 eyes of standardized DSAEK and DMEK that underwent surgery for Fuchs corneal dystrophy at our center. We excluded patients with prior ocular surgery other than cataract surgery to limit confounding variables. METHODS Best spectacle-corrected visual acuity (BSCVA; in logarithm of the minimal angle of resolution [logMAR] units) was obtained and specular microscopy of donor corneal tissue was performed before surgery. Postoperative complications, BSCVA, and the percent of endothelial cell loss (ECL) recorded at 6 months were compared with the Student t test. Patients with pre-existing ocular comorbidity that impacted visual potential such as macular degeneration, amblyopia, advanced glaucoma, and other optic neuropathies were excluded from the analysis of visual acuity, but were included for the outcomes of complications and ECL. MAIN OUTCOME MEASURES Visual acuity improvement, ECL 6 months after surgery, postoperative complications, iatrogenic primary graft failure (IPGF), and rebubbling. RESULTS Of the 200 eyes, 62 DSAEK eyes and 70 DMEK eyes had 6-month BSCVA available and no vision-limiting comorbidities. Mean BSCVA increased from 0.41±0.19 logMAR and 0.27±0.11 logMAR before surgery to 0.20±0.13 logMAR and 0.11±0.13 logMAR 6 months after DSAEK and DMEK, respectively (P<0.001). Seventy-one DSAEK eyes and 70 DMEK eyes had 6-month ECL data available: ECL was 25.9±14.0% after DSAEK and 27.9±16.0% after DMEK (P=0.38). There were no IPGFs in the DSAEK cohort and there were 4 of 100 IPGFs after DMEK (P=0.12). Rebubbling was performed in 2 of 100 eyes after DSAEK and in 6 of 100 eyes after DMEK (P=0.28). CONCLUSIONS Compared with DSAEK, DMEK provided better visual recovery and comparable 6-month ECL. The DMEK group had a higher, although not statistically significant, percentage of rebubbling procedures and IPGFs.

Standardized DMEK Technique: Reducing Complications Using Prestripped Tissue, Novel Glass Injector, and Sulfur Hexafluoride (SF6) Gas.

Purpose: To report low complication rates in Descemet membrane endothelial keratoplasty (DMEK) using sulfur hexafluoride (SF6) gas, a novel glass injector, and donor tissue prestripped by an eye bank technician. Methods: A standardized technique of DMEK was performed in 80 consecutive Fuchs corneal dystrophy cases using technician-prestripped tissue, a novel glass injector, a modified Yoeruek tap technique, and an SF6 gas (20% concentration) bubble for prolonged tissue support. Twenty-five donors were premarked with an “S” stamp for intraoperative orientation. Surgery was performed by 2 experienced DMEK surgeons and 2 inexperienced cornea fellows. Complications were recorded, and the percent endothelial cell loss was calculated at 6 months postoperatively. Results: There were 5 cases that received an air bubble injection postoperatively (6% rebubble rate). There were 6 grafts that immediately failed, 2 because of excessive surgical trauma, and 4 because of upside-down graft placement documented by optical coherence tomography. None of the 25 cases with an S stamp failed. Recipient corneas cleared quickly with no clinical evidence of toxicity from the SF6 gas bubble, and the grafts experienced a mean endothelial cell loss of 27% at 6 months. Conclusions: Tissue prestripped by an eye bank technician can be safely used for DMEK surgery. SF6 gas for prolonged tissue support may reduce the rebubble rate in DMEK, with no apparent acute toxic effect. An unrecognized upside-down graft was the primary cause of graft failure in this series. Upside-down grafts may be eliminated by the use of donor tissue premarked by the eye bank with an S orientation stamp.

Stamping an S on DMEK Donor Tissue to Prevent Upside-Down Grafts: Laboratory Validation and Detailed Preparation Technique Description.

Purpose: To report endothelial cell loss (ECL) caused by a novel S-stamp preparation technique for Descemet membrane endothelial keratoplasty (DMEK). Methods: Six cadaveric human corneas were prepared for DMEK transplantation using a single standardized technique, including the application of a dry ink gentian violet S-stamp to the stromal side of Descemet membrane. Endothelial cell death was evaluated and quantified using computerized analysis of vital dye staining. Results: ECL caused by the S-stamp was 0.6% (range 0.1%–1.0%), which comprised less than one-tenth of the total ECL caused by our preparation of the DMEK graft from the start to finish, including recovery, prestripping, S-stamping, and trephination (13.7% total ECL, range 9.9%–17.6%). Conclusions: Our novel S-stamp donor tissue preparation technique is intuitive to learn and holds the promise of preventing iatrogenic primary graft failure due to upside-down grafts without causing unacceptable increases in ECL.

Relationship Between Tissue Unscrolling Time and Endothelial Cell Loss.

Purpose: To assess the relationship between intraoperative unscrolling time of the donor Descemet membrane endothelial keratoplasty (DMEK) tissue and 6-month postoperative endothelial cell loss (ECL), and to determine whether donor age, scroll tightness, and the presence of an S stamp are related to unscrolling time. Methods: Ninety-three consecutive uncomplicated DMEK surgeries performed on eyes with Fuchs endothelial dystrophy using our standardized technique (ie, prestripped tissue with or without a premarked S stamp from our eye bank, overstripping the recipient, Straiko glass injector, no-touch tap technique, and bubble of 20% SF6 gas) were evaluated. Intraoperative unscrolling times and 6-month endothelial cell densities were measured and analyzed. Results: Sixty-nine cases comprised the study cohort. The median unscrolling time was 4 minutes (range: 0.8–17.5 minutes), and the median ECL was 26.9% (range: −4.3% to 80.0%). There was no relationship between unscrolling time and ECL at 6 months by the Pearson correlation coefficient (r = −0.02, P = 0.89). Younger donor age, tighter scrolls, and absence of an S stamp had no correlation with longer unscrolling times (all P > 0.05). Only 2 of 4 cases of iatrogenic primary graft failure had unscrolling times available for analysis; in this limited sample, there was no association between iatrogenic primary graft failure and unscrolling time. Conclusions: Once the DMEK tissue is safely in the anterior chamber, surgeons need not rush the “DMEK dance” because longer unscrolling times may not endanger the endothelium.

Descemet Membrane Endothelial Keratoplasty (DMEK) Tissue Preparation: A Donor Diabetes Mellitus Categorical Risk Stratification Scale for Assessing Tissue Suitability and Reducing Tissue Loss.

Purpose: This study assessed a novel diabetes mellitus (DM) rating scale in relation to its utility in reducing Descemet membrane endothelial keratoplasty (DMEK) tissue preparation failure. Methods: A 5-point DM rating scale was defined, in which 1 demonstrated relatively good health associated with DM and 5 represented comorbidities associated with DM. A chart review from consecutive donors who had at least 1 tissue prepared for DMEK was performed. Using the donor profile, the first tissue processed from each donor was categorized according to the DM severity and if the tissue passed or failed the DMEK preparation. Failure rates per rating group were evaluated using logistic regression and odds of preparation failure. Results: A total of 125 tissues prepared for DMEK were categorized based on the defined DM rating scale. Of these, 9 tissues were rated 1 (11.1% failure), 25 were rated 2 (0% failure), 31 were rated 3 (6.5% failure), 24 were rated 4 (16.7% failure), and 36 were rated 5 (30.6% failure). The odds ratios were significant for tissues rated as 5 and 3 (P < 0.05). No other rating categories were found to influence the odds of failure. A &khgr;2 test comparing categories of low risk (1–3) and high risk (4–5) was also performed (P = 0.001). Conclusions: The DM rating scale does seem to stratify the risk of preparation failure associated with the severity of DM and associated comorbidities. Inclusion of some diabetic donors for the preparation of DMEK grafts may be warranted given proper screening of the donor history and application of the rating scale.

Corneal Astigmatism Stability in Descemet Membrane Endothelial Keratoplasty for Fuchs Corneal Dystrophy.

Purpose: To calculate the magnitude and angle of the shift in corneal astigmatism associated with Descemet membrane endothelial keratoplasty (DMEK) surgery to determine the feasibility of concurrent astigmatism correction at the time of DMEK triple procedures. Design: Retrospective study. Methods: Forty-seven eyes that previously underwent the DMEK procedure for Fuchs endothelial corneal dystrophy and that had more than 1.0 diopter (D) of front corneal astigmatism preoperatively were identified. All DMEK surgeries used a clear corneal temporal incision of 3.2 mm. Surgically induced astigmatism (SIA) was evaluated 6 months postsurgery with vector analysis using Scheimpflug image reading. Results: We did not find a difference between pre- and postoperative magnitude of front astigmatism (P = 0.88; paired t test). The magnitude of the SIA front surface was 0.77 ± 0.63 D (range, 0.10–3.14 D). The centroid vector of the SIA front surface was 0.14 at 89.3°. A hyperopic corneal power shift was noted in both the front surface by 0.26 ± 0.74 D (range, 0.45–3.05 D) (P = 0.018; paired t test) and back surface by 0.56 ± 0.55 D (range, 0.25–2.40 D) (P < 0.01; paired t test). Conclusions: DMEK surgery induces minimal amounts of corneal astigmatism that is a with-the-rule shift associated with a temporal clear corneal incision. The stability of these data from preop to postop supports the plausibility of incorporating astigmatism correction with the cautious use of toric intraocular lenses for patients with Fuchs corneal dystrophy and cataract.

Astigmatism Correction With Toric Intraocular Lenses in Descemet Membrane Endothelial Keratoplasty Triple Procedures.

Purpose: To report the clinical efficacy of astigmatism correction with toric intraocular lenses (IOLs) in patients undergoing the Descemet membrane endothelial keratoplasty (DMEK) triple procedure and to evaluate the accuracy of the correction. Methods: Fifteen eyes of 10 patients who received cataract extraction, toric IOL placement, and DMEK surgery for Fuchs corneal dystrophy and cataracts were evaluated. The cylinder power of toric IOLs was determined by an online toric calculator with keratoscopy measurements obtained using Scheimpflug corneal imaging. Prediction errors were assessed as a difference vector between the anticipated minus postoperative residual astigmatism. Results: At 10.1 ± 4.9 months postoperatively, 8/13 (61.5%) of eyes achieved uncorrected distance visual acuity better than 20/40. Mean best spectacle-corrected distance visual acuity (logMAR) improved from 0.21 ± 0.15 preoperatively to 0.08 ± 0.12 postoperatively (P < 0.01). The magnitude of refractive astigmatism was also significantly decreased from 2.23 ± 1.10 D (range 0.75–4.25 D) preoperatively to 0.87 ± 0.75 D (range 0.00–3.00 D) postoperatively (P < 0.01). In 1 eye with rotational misalignment by 43 degrees, we found no improvement of astigmatism. The prediction error of astigmatism at the corneal plane was 0.77 ± 0.54 D (range 0.10–1.77 D). Four eyes with preoperative “with-the-rule” corneal astigmatism had postoperative “against-the-rule” refractive astigmatism. Conclusions: For patients with Fuchs corneal dystrophy and cataracts, use of toric IOLs might be a valuable option in triple DMEK surgery. Additionally, care should be taken to prevent excessive IOL rotation.

DMEK Graft Preparation: Eye Bank Perspective and Risk Factors for Preparation Failure

DMEK is the fastest-growing corneal transplant procedure for treating endothelial dysfunction in the USA. DMEK grafts prepared by eye banks save surgeons time in the operating room, eliminate failed tissue preparations performed by the surgeon, and provide post-processing quality assurance. While these benefits lower the barrier to DMEK adoption, establishment of a superb DMEK graft program is no easy task. A DMEK program requires careful consideration of tissue selection, staff training, post-processing tissue evaluation, and, perhaps most importantly, selection of a preparation technique.

Descemet Membrane Endothelial Keratoplasty (DMEK) Surgery with a Standardized Technique

The goal of this chapter is to provide the DMEK surgeon with a toolkit of techniques to prevent and manage perioperative complications by unpacking the procedure step by step. The chapter is organized into preoperative and intraoperative considerations. Specific techniques for unscrolling the graft, because there are so many, are not reviewed in great detail, either. A summary table is provided for quick reference (Table 9.1). A video tutorial is also available and is indispensable to mastering the concepts set forth in this chapter.

Donor endothelial cell density measurements do not change immediately after DMEK preparation

Purpose: To evaluate a single eye banks measurement of endothelial cell density (ECD) of Descemet membrane endothelial keratoplasty (DMEK) grafts before and after preparation using 2 separate counting methods. Methods: A series of 60 donor tissues were prepared for DMEK surgery. One to 4 specular images of the central endothelium were taken both before and after preparation, and ECDs were evaluated for a total of 345 unique images. Images were then masked and provided to the Cornea Image Analysis Reading Center (CIARC) for independent analysis. Results: Before preparation, average eye bank-determined ECD with the center method was 2678 ± 259 cells/mm2 and was 2599 ± 280 cells/mm2 CIARC-determined by the variable frame method (P < 0.001, n = 176). After preparation, eye bank-determined ECD was 2719 ± 265 cells/mm2 and CIARC-determined ECD was 2615 ± 344 cells/mm2 (P < 0.001, n = 169). The difference in ECD before and after DMEK preparation was not found to be statistically significant when evaluated using either analysis method (P = 0.19; P = 0.64) before and after preparation, respectively. Conclusions: Although the absolute ECD value may differ by the analysis method statistically, pre- and post-DMEK preparation ECDs did not significantly change by either analysis method. Other methods such as vital staining to assess tissue damage after preparation in conjunction with specular microscopy are suggested.