To log in, click the teal "Login" button in the upper right-hand corner of this page. If you are logged in but still do not have access, please check your 2026 Annual Meeting registration.
Moderator
Yousuf Khalifa, MD, FACS
Panelists
Masih U Ahmed, MD, ABO
Viewing Papers
Expand a paper title to the right to view the paper abstract and authors. Use the video link to jump to that poster in the session.
Presenting Author
Allison J. Bernstein, MD
Co-Authors
Julia Talajic (MD, FRCSC), Siddharth Nath (MD, PhD), Johanna Choremis (MD, FRCSC), Louis Vaudoisey (MD), Samir Jabbour (MD, FRCSC)
Purpose
To assess visual, refractive, and safety outcomes of toric intraocular lenses (IOL) in patients with Fuchs endothelial corneal dystrophy (FECD) undergoing combined or staged Descemet membrane endothelial keratoplasty (DMEK).
Methods
This retrospective case series included 31 eyes of 26 patients with stage I FECD, astigmatism, and cataract undergoing DMEK in a staged or combined approach with implantation of a toric IOL. Patients were identified from an existing database encompassing three affiliated sites under the Universit� de Montr�al in Montreal, QC. Postoperative corrected and uncorrected distance visual acuities (CDVA/UDVA), spherical equivalent (SE), residual cylinder, endothelial cell loss (ECL), and complications were documented and compared between staged and combined groups pre- and postoperatively.
Results
CDVA improved significantly from 0.34 logMAR preoperatively to 0.07logMAR postoperatively in both DMEK-cataract surgery groups (p<0.05). Statistically significant improvements were noted from preoperative CDVA to post-operative UDVA, from 0.34logMAR to 0.11logMAR (p<0.05) in both groups. Mean residual SE was -0.27D, with no significant difference between staged and combined, but a trend towards lower residual SE in the combined cohort. Residual cylinder was 0.94D and 0.46D in the combined and staged groups. There was more endothelial cell loss (ECL) in the combined group, but not significantly. Fewer complications occurred in the staged group, though overall complication rate was low.
Conclusion
Toric IOLs are safe and effective in managing astigmatism in the setting of stage I FECD treated with DMEK. Staged DMEK-cataract surgery is associated with fewer complications.
Presenting Author
Feyza D. Isik, MD
Co-Authors
Emine Esra Karaca (MD), Ozlem Evren Kemer (MD), Kas?m Öztoprak (PhD)
Purpose
To develop temporalspecific artificial intelligence models predicting clinical outcomes after DMEK using IVCM and clinical parameters, validate their performance against expert assessment, and quantify systematic divergence between algorithmic predictions and physician judgment to establish the critical role of experts in posttransplant management.
Methods
Retrospective analysis of 75 DMEK recipients(2019-2023) with IVCM at months 3,6,12.Three approaches developed:HealthyRefNetRF,CNN (EfficientNetV2B3)trained on 3,100 control images;ExpertVisionRF, CNN trained on 3,100 DMEK images;PureExpertRF,expert binary assessment without image processing.Random Forest integrated CNN features (models 1-2) or binary ratings (model 3) with clinical parameters (demographics, indication, thickness, acuity, comorbidities,etc.).Concordance between predictions and surgeon-determined outcomes(favorable/unfavorable) assessed via AUC-ROC, Cohen's ?,McNemar test,multivariate regression and 5-fold cross-validation.
Results
Models showed temporal specificity.ExpertVisionRF dominated at month 3(AUC:0.894,?:0.78),HealthyRefNetRF at month 6 (sensitivity:95.7%,AUC:0.806),PureExpertRF showed highest clinical concordance at month 12(?:0.595).Systematic model-physician divergence peaked at month 6(p<0.002,McNemar),17.3% of patients were misclassified with contradictory predictions for identical cases(9.3%).Preoperative corneal thickness ?690?m predicted inability to obtain early postoperative specular microscopy images(specificity 85.2%),establishing the 1st evidence-based threshold for imaging modality.Intraoperative epithelial debridement found as the independent predictor of poor 1 year prognosis(OR:3.90,p=0.024).
Conclusion
Phase specific algorithms revealed DMEK's temporal complexity;early pathology detection,transitional monitoring and late stability required distinct computational approaches.The 690?m imaging threshold provides clinical tool.No algorithmic blind spots validate methodology.Medical AI must evolve from static classifiers to phaseaware dynamic systems.
Presenting Author
Zeynep Hande Kaya, MD
Co-Authors
Emine Esra Karaca (MD), Ozlem Evren Kemer (MD)
Purpose
To assess the safety and efficacy of DMEK in treating post-PK endothelial decompensation,evaluate the prevalence of cytomegalovirus(CMV) and its clinical implications,identify predictors for re-DMEK and rebubbling, and establish risk stratification models based on viral,anatomical,and surgical factors influencing graft survival and visual outcomes.
Methods
A series of 30 eyes underwent DMEK for failed PKP(2020-2025). Eligilibity criteria: endothelial decompensation >1 year post-PKP with preserved central stroma. Pre-op evaluation included viral PCR, pachymetry, and anterior segment imaging. A standardized DMEK was performed, with graft size based on PKP diameter. Outcomes will be assessed at 1, 3, 6,and 12 months which are visual acuity,clarity,complication. The analysis will consider various factors: PKP age, rejection, viral status (CMV/HSV), glaucoma,rebubbling, and re-DMEK. Statistical analysis will involve: survival curves and regression models.Antivirals were given per PCR results. Success defined as graft clarity without reintervention.
Results
Thirty patients underwent PK-DMEK with a mean PK-DMEK interval of 61.27±54.68 months. Postoperative re-bubbling occurred in 30% of cases (n=9). No significant difference was observed between re-bubbling and tamponade type (p=0.29), glaucoma history(p=0.25),or aqueous humor PCR results (p=0.75). Mean BCVA improved significantly from 1.71±0.66 LogMAR preoperatively to 0.99±0.45 at month 12. Wilcoxon test confirmed significant improvement at months 1 (p=0.01), 3 (p<0.01),6 (p=0.008),and 12 (p=0.01). At one year, mean corneal thickness was 534.00±56.38 μm and mean ECD was 1559.00±864.84 cells/mm². CMV prevalence was 25%,and primary graft survival was 90%.
Conclusion
DMEK is an effective,less invasive technique treating endothelial failure after PK.Although limited sample size and follow-up loss restricted establishing the statistical predictor correlations,a 90% survival rate is promising.Visual improvement and deturgescence were significant.Viral PCR and risk assessment are vital for optimizing complex cases.
Presenting Author
Dilara Altun, MD
Co-Authors
Ozlem Evren Kemer (MD), Emine Esra Karaca (MD)
Purpose
To evaluate clinical outcomes and risk factors after Descemet membrane endothelial keratoplasty (DMEK) in vitrectomized eyes.
Methods
In this retrospective review of 480 DMEK procedures performed at our clinic between January 2018 and January 2025, 58 patients with a history of vitrectomy and regular follow-up were included. The mean follow-up was 19 months. Patients' prior ocular surgeries and clinical characteristics were recorded. Examinations were performed preoperatively and at postoperative week 1 and months 1, 3, 6, and 12. Best-corrected visual acuity (BCVA, logMAR), central corneal thickness (CCT), intraocular pressure (IOP), and endothelial cell density (ECD) were assessed.
Results
The mean preoperative BCVA was 2.06±0.62 logMAR,improving to 1.15 ± 0.88 at month 1(p<0.001).Mean preoperative CCT was 755.8 ± 110.3 µm and decreased to 571.2 ± 80.7 µm at month 1(p<0.001). Mean ECD loss was 42.9±19.4% at month 6 and 52.4±18.3% at month 12.Graft detachment occurred in 16 patients at a mean of 3.2±7.9 days postoperatively.11 patients underwent re-DMEK—three early for persistent non-attachment and eight later for graft failure—at a mean of 278±340 days.Four patients underwent penetrating keratoplasty at a mean of 13±8 days.A longer interval between vitrectomy and DMEK was significantly associated with the development of CME(p=0.024)
Conclusion
In vitrectomized eyes, DMEK is challenging due to loss of positive posterior support from the vitreous. A deep anterior chamber complicates graft unfolding and positioning, potentially prolonging surgery and increasing risks of graft detachment and endothelial cell loss. Nevertheless, favorable visual and anatomical outcomes are achievable.
Presenting Author
Leonard G Heydenrych, FRCOphth
Co-Authors
Eric Abdullayev (MD), Benjamin Lambright (MD)
Purpose
To establish if there is a difference in tapping rates and graft unfolding times when �No-touch' Endo-In DMEK is compared with standard DMEK.
Methods
Real time video (Zeiss) of 41 �no-touch' DMEK surgeries were reviewed retrospectively and compared with 41 real time videos retrospectively when standard DMEK was performed. Tapping rates and graft unfolding times were calculated until the grafts were opened completely, after graft injection into the anterior chamber. A Brunner-Munzel test was used to compare the two groups with each other, after calculating the standard deviation and mean for each category. Each surgery was reviewed and the time required to unfold the graft completely was calculated as well as the amount of taps required to unfold the graft completely.
Results
The mean tapping rate for �no-touch' Endo-In DMEK was 22.6829 taps,(SD=42.4096). The mean tapping rate for standard DMEK was 137.7561 taps,(SD=134.1545). In this sample, �no-touch' Endo-In DMEK required 83.534% fewer taps to unfold the graft.The mean graft unfolding time for �no-touch' Endo-In DMEK was 39.4146 seconds, (SD=51.662). The mean graft unfolding time for standard DMEK was 129.9512 seconds (SD=117.3297). In this sample, 'no-touch' Endo-In DMEK required 69.67% less time to completely unfold the graft.
Conclusion
The Brunner-Munzel test confirmed a statistically significant reduction in tapping rate and time to completely unfold the DMEK graft when �no-touch' Endo-In DMEK was performed. �No-touch' Endo-In DMEK is a less variable and more predictable procedure when compared to standard DMEK.
Presenting Author
Hyeck Soo Son, MD
Co-Authors
Victor Augustin (MD), Maximilian Friedrich (MD), Gerd Auffarth (MD, PhD)
Purpose
To describe the long-term observations of the first human implantation of EndoArt, a synthetic endothelial implant, in a small cohort of patients with corneal edema secondary to endothelial dysfunction.
Methods
This retrospective case series analyzed 5 eyes from 5 patients who underwent EndoArt implantation, with a minimum follow-up of 12 months. Clinical outcomes, including corrected distance visual acuity and central corneal thickness, were evaluated pre- and postoperatively. Rebubbling procedures and any complications were recorded. Histological analysis was performed on 1 explanted corneal button.
Results
The mean follow-up was 38.4 ± 20 months (range: 19-72 months). A reduction in mean central corneal thickness was observed, from 741.2 ± 88.3 ?m preoperatively to 558.4 ± 72.7 ?m postoperatively (P < 0.001). Corrected distance visual acuity improved in 4 of 5 eyes, although not significantly overall (P = 0.19). The mean number of rebubbling procedures was 2.8 ± 1.3 per eye. No serious implant-related complications were observed during the follow-up period. Histological findings showed fibrocellular tissue partially integrating the implant, which may contribute to graft adherence.
Conclusion
This small pilot series of the first human EndoArt implantation observed a significant reduction of corneal edema over time and a favorable safety profile in select patients. Larger, controlled studies are needed to better understand the safety, efficacy, and long-term outcomes of this synthetic implant.
Presenting Author
Mirac S. Kucukoner, MD
Co-Authors
Ozlem Evren Kemer (MD), Emine Esra Karaca (MD)
Purpose
To evaluate risk factors and clinical outcomes of repeat Descemet membrane endothelial keratoplasty (re-DMEK) for failed primary DMEK grafts at a referral center in Ankara, Türkiye.
Methods
A retrospective, interventional, comparative case series involving 590 DMEK surgeries performed by two surgeons at Ankara Bilkent City Hospital between July 2020 and August 2025. Clinical data were analyzed for eyes requiring re-DMEK due to primary graft failure. Patients who underwent DMEK and re-DMEK surgery were analyzed in terms of previous glaucoma history, vitrectomy, uveitis/endotheliitis, and HSV and CMV infection positivity as risk factors.
Results
Of 590 DMEK surgeries 73 eyes were underwent re-DMEK, 57 eyes (78%) underwent one re-DMEK, 13 eyes (18%) had two, 2 eyes (2.7%) had three, and 1 eye (1%) underwent four consecutive re-DMEK procedures. Sixteen eyes had prior vitrectomy, 14 of which were due to complicated cataract surgery. Glaucoma history was present in 17 patients, with 9 having undergone glaucoma surgery prior to DMEK. Rebubbling was performed in 15 eyes post-re-DMEK. Four cases tested positive for CMV PCR, and one case was clinically consistent with HSV infection. Preoperative BCVA was 1.51 ± 0.80, and at 1 year postoperative, BCVA significantly improved to 0.85 ± 0.64 (p < 0.001).
Conclusion
Repeat DMEK is a viable option for managing primary DMEK failure. Although several risk factors exist for graft failure, re-DMEK still offers satisfactory clinical results. Complicated cataract surgery emerged as the primary risk factor for primary DMEK failure, while rebubbling was the most significant risk factor for re-DMEK failure.
Presenting Author
Dilara Altun, MD
Co-Authors
Ozlem Evren Kemer (MD), Emine Esra Karaca (MD)
Purpose
To characterize anterior segment OCT-based interface findings in eyes requiring rebubbling after DMEK, describe their changes over time, and assess associations with clinical risk factors and tamponade characteristics.
Methods
We retrospectively reviewed all DMEK procedures performed at our center between January 2018 and January 2025 (n=480); eyes undergoing ?1 rebubbling comprised the analytic cohort.AS-OCT were processed in ImageJ using a standardized densitometry macro.The cornea was partitioned into concentric 500-µm annular zones.For each zone—and for clinically detached interfaces—we quantified stromal mean gray value (MGV) and the anterior-to-posterior stromal MGV ratio. Interface MGV/irregularity were compared between detached vs attached zones and over time and correlated with predefined risk factors to guide the need and timing of re-intervention.
Results
TBD
Conclusion
TBD
Presenting Author
Ansu A John, MBBS, MS
Co-Authors
Jacob Fondriest (MD), Ellen Koo (MD)
Purpose
To compare Endothelium-in (Endo-in) Descemet Membrane Endothelium Keratoplasty (DMEK) with DMEK performed using a glass injector system, focusing on operative time, visual acuity, pachymetry, and safety outcomes to guide surgical planning.
Methods
A retrospective study of 76 eyes undergoing DMEK over one year was performed, including standalone DMEK and combined phacoemulsification/DMEK performed entirely by fellows. Exclusion criteria included prior herpes simplex virus/varicella zoster virus, uveitis, or allogeneic rejection. Outcomes were operative time, best corrected visual acuity (BCVA) at 3 and 6 months, pachymetry at 3 and 6 months, rebubbling rate, and graft failure. Endo-in cases were compared with those using a traditional glass injector system (Jones tube injector).
Results
Endo-in DMEK techniques showed significantly shorter intraoperative time compared with Jones tube glass injector system (p<0.05). At 6 months, BCVA improved by +14 letters (endo-in) vs +3 letters (injector, p=0.88). Pachymetry decreased 96 µm (endo-in) vs 225 µm (injector, p=0.18). Rebubbling occurred in 16% of endo-in eyes (8/49) vs 19% with injector (5/27). No primary graft failures occurred.
Conclusion
Endothelium-in DMEK reduced operative time compared with Jones tube glass injector system and showed a trend toward improved visual acuity with comparable safety and pachymetry outcomes. The technique appears safe and feasible, but larger prospective trials are needed to confirm clinical advantages.
Presenting Author
Uma Yogesh Thigale, MBBS, MS
Co-Authors
Pravin Vaddavalli (MD)
Purpose
To emphasize the importance of availability of advanced ocular health care to rural people at Secondary Health Care unit which are in proximity of their place of residence and work.
Methods
This is a case series of patients with Pseudophakic Bullous Keratopathy who underwent DMEK at a Secondary Health Care (SHC) level eye unit. Three eyes (three patients; mean age, 65 years) with endothelial decompensation as a complication of cataract surgery were included. Diagnosis, pre-operative ophthalmic and systemic investigations were done at SHC level, minimizing the need for referral to Tertiary Health Care (THC) unit. All surgeries were performed by a single cornea surgeon from August 2017 to August 2018 at SHC level unit. Corneal tissue was collected and carried from THC eye bank by the surgeon himself , ensuring maintenance of cold-chain and reducing the transportation cost
Results
Post-operative visual outcomes were measured at one, three, six and twelve months post-operatively. Out of three, two eyes showed BCVA of 20/20 at 6 months post DMEK . One patient had BCVA of 20/80p at 6 months and sub-optimal vision was attributed to decreased vision post vitreo-retinal surgery (which was performed at THC unit and followed up at SHC unit); however the patient had a BCVA of 20/40 at 12 months follow-up. None of the patients had any episode of post-operative complication related to DMEK.Follow-up examination was done by trained Cornea fellows posted at SHC unit. The follow-up care and investigations were monitored by the operating surgeon by Teleophthalmology from THC unit.
Conclusion
Patients do not follow up at THC due to the inconvenience & unaffordability of travel,loss of work hours & wages.Specialised care post corneal transplantation can be provided to patients at SHC unit ensuring better follow-ups,cost-effectiveness, reduced loss of work-hours for patients & earlier management of graft & complications post-operatively.
