Rapid corneal endothelial cell loss following <scp>PreserFlo MicroShunt</scp> implantation

PreserFlo MicroShunt (PMS) (Santen Pharmaceutical Co. Ltd., Osaka, Japan) is an ab-externo drainage device offering a safer alternative to trabeculectomy for lowering IOP, and is becoming widespread (Panarelli et al., 2024). The long-term safety of the corneal endothelium remains a concern. Although initial reports suggested a favourable safety profile with minimal endothelial cell loss (ECL), recent literature has described ECL following PMS implantation (Damgaard et al., 2025; Ibarz-Barberá et al., 2022; Iwasaki et al., 2025; Watanabe et al., 2025), indicating that the full safety profile and the mechanisms driving endothelial damage remain poorly understood. Between January 2023 and October 2024, 276 eyes of 229 patients underwent PMS implantation at Hiroshima University Hospital. We identified six eyes of six patients (one female and five males) (incidence: 2.2%) that exhibited rapid ECL necessitating device removal. The mean age was 75 ± 8.8 years. Five eyes had primary open-angle glaucoma, and one had exfoliation glaucoma. The mean preoperative intraocular pressure (IOP) and corneal endothelial cell density (ECD) were 20.0 ± 5.6 mmHg (range, 12–28 mmHg) and 2189 ± 592 cells/mm2 (range, 1457–2840 cells/mm2), respectively. Three eyes had a history of intraocular surgery: phacovitrectomy, phacotrabeculotomy and cataract surgery. PMS implantation was performed as a solo procedure in three cases and combined with cataract surgery in others. No intraoperative complications were recorded. Figure 1a shows anterior-segment optical coherence tomography (AS-OCT) for each eye at 6 months post-operation. The mean tube-cornea distance (TCD) and tube-cornea angle (TCA) were 467 ± 145 μm (range, 305–692 μm) and 20.7 ± 2.7° (range, 17.4–25.2°). Figure 1b shows the ECD course from preoperative to PMS removal. At 3, 6 and 12 months postoperatively, the mean ECD were 1534 ± 309, 1231 ± 167, and 995 ± 243 cells/mm2, respectively, and the mean rate of ECD decrease was 27.9%, 40.4% and 50.2%, respectively. Immediately before PMS removal, each had 849 ± 249 cells/mm2 and 60.4%. At the time of PMS removal, trabeculectomy was concurrently performed in five eyes, and Ahmed glaucoma valve implantation into the ciliary sulcus in one eye. For the five eyes in which postoperative ECD measurements were obtained, the progressive decline in ECD halted after PMS removal. This outcome aligns with previous case reports (Iwasaki et al., 2025; Watanabe et al., 2025). This finding strongly suggests a relationship between PMS and ECL and indicates the effectiveness of surgical intervention in preventing irreversible corneal decompensation. Although a short TCD (<200 μm) has been reported as a risk factor (Ibarz-Barberá et al., 2022), our cases showed rapid ECL despite being within the safe range. However, AS-OCT indicated the tube angulation towards the cornea in several cases. Unfortunately, the insertion site of the tube was not confirmed using gonioscopy. Therefore, it cannot be ruled out that direct physical contact of the tube is the cause of corneal damage. The six cases exhibited a substantial reduction in ECD, ranging from 21% to 61%, within one year. This rate of loss is markedly greater than the average annual reduction of 5.2%–7.4% reported in an extensive cohort study (Ibarz-Barberá et al., 2022; Panarelli et al., 2024). Such a significant discrepancy implies that factors beyond those currently established, or a confluence of multiple factors, contributed to this accelerated ECL in our patients. In conclusion, severe ECL can be a complication after PMS implantation. As it remains unclear which specific patient profiles are susceptible to this rapid ECL, diligent, serial monitoring of ECD is essential. Our findings demonstrate that progressive ECL can be consistently halted by PMS removal and conversion to another glaucoma surgery. Conceptualization: K. Hirooka; Data acquisition: S.D.; data analysis and interpretation: S.D., K. Hirooka.; figure and table preparation: S.D.; writing of the main manuscript text: S.D.; critical review and revision of the manuscript: K. Hirooka, K. Hirose., T.B., N.O., K.T., A.F., T.C., H.S. We would like to thank Editage (www.editage.jp) for English language editing. The authors declare that no funds, grants or other support were received during the preparation of this manuscript. The University of Hiroshima Institutional Review Board approved this retrospective study (approval no. E-2436). Standard written informed consent for surgery was obtained from all patients. Restrictions apply to the availability of these data, and they are, therefore, not publicly available. The data underlying this study are available upon reasonable request from the corresponding author (K.H., [email protected]) with the permission of Hiroshima University Hospital.