• Cornea/External Disease, Refractive Mgmt/Intervention

    In this retrospective cohort study, researchers assessed how corneal shape predicts hyperopia after triple Descemet membrane endothelial keratoplasty (DMEK).

    Study design

    This cohort included 112 eyes from 101 patients with Fuchs endothelial corneal dystrophy who underwent uncomplicated triple DMEK at a single center. These patients had Scheimpflung exams prior to the procedure. The arithmetic refractive predictive error was calculated based on stable postoperative refraction minus predicted refraction from biometry. Risk ratios were categorized into hyperopia (>+0.5 D), myopia (>0.5 D) or emmetropia (±0.5 D). The IOL types used were CT Asphina, SA60WF, SN60WF and SA60AT.


    Overall, the arithmetic refractive error was 0.34 D. Approximately 46% of eyes had a hyperopic arithmetic error and 38% had an emmetropic error; myopic errors were less common (16%).

    Approximately 60% of eyes with an oblate cornea (positive posterior Q value) were hyperopic compared with 30% of eyes with prolate corneas. Oblate corneas had a 3-fold higher risk of hyperopic shift after triple DMEK compared with spherical or prolate corneas (P=0.011). The mean prediction error was 0.5 D higher in eyes with positive Q values (P=0.002), independent of corneal thickness.

    Hyperopic errors occurred more frequently in corneas that were more centrally flat, possibly due to edema changes.


    The retrospective nature of the study limited the cohort to patients who had preoperative topography. It is unexpected that the authors reported high rates of a hyperopic predictive error. Astigmatism measurements would have been an interesting addition.

    Clinical significance

    Surgeons looking for the optimal refractive outcome following DMEK with cataract surgery and IOL implantation should look at preoperative topography and aim for more myopia in eyes that have oblate posterior cornea.